Scholarship in Progress

This Article uses a massive collection of data to document, for the first time, the interplay between the substantive subject areas, the intensity, and the procedural complexity of federal civil litigation. The results indicate that like substance, intensity and complexity are pivotal features of litigation. These findings suggest that what might be termed “our uniformity”—the de jure uniform applicability of the Federal Rules of Civil Procedure and other procedural law—should be understood as a broader phenomenon than the traditional focus on transsubstantivity. We have both transintensity and transcomplexity as well—and in a word, we have transintexity.

The Article documents extensive variation in measures of intensity and complexity among cases sharing the same broad substantive legal subject matter, such that intensity and complexity may vary more importantly within than across substantive areas. For example, it is true that patent cases have comparatively high average intensity as measured by their numbers of docket entries, but there are also plenty of patent cases that terminate without having enormous docket activity. It is also true that patent cases are much more likely than, say, contracts cases to be extremely intense, but  because there are so many more contracts than patent cases, my data have nearly twice as many highly intense contracts cases as highly intense patent cases. The Article’s twin key conclusions are, thus: (1) substance is transintex—e.g., there are both high- and low-intexity patent cases—and (2) intexity is transsubstantive—e.g., there are highly intex cases in most substantive areas of litigation.

The Article closes by treading gingerly into normative waters. One unavoidable consequence of our uniformity with respect to formal procedural rules is that judges exercise enormous case-level discretion. Further, in part due to legislative forays in the securities and patent arenas, we have some degree of substance-specificity in our procedure. The Article suggests considering formal procedural tracking as an alternative to each of these approaches. That approach, which has been suggested in the past and is used in some states and other countries, might allow us to break out of some of our ossified debates about matters such as the rancor set off by the pleading revolution the Supreme Court effected a decade and a half ago in Twombly and Iqbal. If the plausibility standard is here to stay because of its role in limiting intense and complex litigation, perhaps it could be revisited in simpler cases that don’t pose the challenges the Supreme Court first flagged in Twombly. Whether this is possible depends on our capacity to identify cases’ likely intexity early in the litigation life cycle, which is a topic beyond the scope of this Article. Still, we ought to consider the possibility of a certification process for intense, complex cases, like the one we have for class actions, so that procedure might be adjusted where doing so makes sense.

The Article’s contributions include its use of docket-level data on more than 500,000 cases whose dockets could be followed for at least 7 years. In addition, the Article offers a novel approach to measuring procedural complexity by showing how links between entries in docket reports may be viewed as mathematical networks. Finally, the Article’s empirical analysis is conducted entirely using simple descriptive statistics, so that no understanding of multivariate estimation or other techniques is necessary.

This chapter starts with the observation that many statutes may be regarded as functions linking a relevant object—e.g., a person or an employment action—can be linked to legal statuses, such as of-drinking-age or unlawful-employment-practice. That function may often be separated into a qualification function that links the relevant object to a qualification condition, and a legal status function that links qualification conditions to legal status. Qualification functions change over time even when a statute’s own text does not: Consider what happens when the legal drinking age is defined in terms of the age of majority, which is defined elsewhere in the code and changes. Consequently, it is impossible for a theory of statutory interpretation to require interpretation of textually-constant statutes so that both (1) results are unchanged over time, and (2) qualification conditions are interpreted the same way across areas of the law at a given time.

I call that the dynamic dilemma. One consequence of the dilemma is that abstract theories of interpretation, such as textualism or originalism, can’t resolve key questions except via ipse dixit. Which disuniformity one chooses has to be justified on substantive grounds. I illustrate this fact using the effects of the 1920 ratification of the 19th Amendment on state jury-service statutes. Statutes enacted before 1920 made only “qualified electors” eligible for jury service. Women couldn’t be jurors before 1920 due to laws excluding them from the franchise—but what about afterward? States had to choose between allowing textually-constant jury-service statutes to give different results over time or to treat “qualified elector” differently after 1920 than its treatment in other areas of the law. The abstract “rules” of statutory interpretation that the Pennsylvania and Illinois Supreme Courts used to answer these questions are a classic of the canon/countercanon form, illustrating the futility of responding to the dynamic dilemma that way.

I also discuss the Supreme Court’s Bostock v. Clayton County decision, which determined that employment discrimination on the basis of sexual orientation and transgender status violates Title VII’s prohibition of discrimination “because of” an “individual’s … sex.” Drawing on—perhaps simply applying—arguments by Eskridge, Slocum, and Gries in their 2021 article, The Meaning of Sex, I suggest Title VII might have been interpreted differently in 1964 and 2021 due to the combined effects of evolving social facts and the role of statutory interpretation’s absurdity doctrine in scoping the word sex in 1964. Perhaps Justice Alito’s dissenting critique—that Bostock was wrong because no one in 1964 would have read Title VII as the majority did—is both right about 1964 and wrong about 2021. This argument, rooted in the dynamics of disuniformity, contrasts with influential analyses that would resolve Bostock on grounds that would yield the same answer in 1964 and 2021.

Finally, I argue the dynamic dilemma is a special case of a broader one involving disuniformity and the multiple-dimensionality of law. Using the example of how law applies in state and federal courts under the Erie doctrine, I point out that in a legal system that allows heterogeneity within more than one dimension of applicable law, legal rules must sometimes vary either within or across dimensions.