The Belt-and-Suspenders Canon

I.     Introduction

We aim here to christen a canon into the doctrines of statutory interpretation. We propose that a “belt-and-suspenders” canon, invoked with some frequency during the legislative drafting process, should be recognized by judges and scholars as presumptively probative and at times controlling. Two recent high profile cases—Yates v. United States1 and Hively v. Ivy Tech2—consider the possibility that the relevant legislatures whose work product was at issue wrote their statutes with features that were deliberatively duplicative, redundant, and/or reinforcing rather than perfectly parsimonious.3 This appreciation for the realities associated with legislative drafting bears a family resemblance to an older canon recently rediscovered by some prominent jurists: ex abundanti cautela,4 translated as “in an abundance of caution.” Historically more than recently, courts have recognized that legislatures can draft statutes to be abundantly cautious rather than to be supremely concise.5 Yet the idea that legislatures use belts and suspenders is ultimately in tension with the hoary canon that statutes ought to be presumed to contain no superfluities, a canon widely recognized even if subject to frequent critique.6

That an individual canon can sometimes be met with a counter-canon undermining its persuasive force is hardly a new discovery. Karl Llewellyn’s “thrust but parry” elegantly mapped this dynamic, though he did not anticipate that “belt-and-suspenders” would do battle with “anti-surplusage.”7 A recent study of “dueling canons” also does not take note of this contradictory couplet of canons.8 Yet the tensions involving the two canons reveal a deeper strain between the legislative and judicial branches.

This Essay, by illustrating and unpacking the popularity for legislatures of a “belt-and-suspenders” approach to drafting, helps explain the legislative dynamics that give rise to reasonable redundancy. After identifying this legislative rationality, the Essay then begins the project of helping courts decide when to apply what we are calling the “belt-and-suspenders” canon, and when it might make sense to fall back on their rule against superfluities. This analytical project of differentiating the relative appropriateness of different canons for different contexts has value both for judges whose interpretive approach to statutes focuses exclusively upon textual meaning and for judges whose approach also emphasizes underlying legislative purpose and intent.

In its various iterations,9 the rule against superfluities has received talismanic judicial endorsement.10 Justice Scalia (with Bryan Garner) devotes six pages of his treatise to justifying the canon in historical and pragmatic terms.11 Legislatures, however, have been less enamored. A recent study by Abbe Gluck and Lisa Bressman of more than 130 congressional staffers indicates that although almost two-thirds of the staffers knew the rule against superfluities by some version of its name, well over half of respondents said the rule should apply in legislative drafting rarely or, at most, only sometimes.12 By contrast, staffers reported more robust drafter reliance on other language canons such as noscitur a sociis and expressio unius,13 which actually have less support in the courts than does the rule against superfluities.14 And in a study of state laws that codify canons of statutory interpretation, only ten states reduced the anti-surplusage canon into their codes.15

By contrast, Scalia and Garner identify the “belt-and-suspenders approach” as “ill-conceived but lamentably common,”16 and Justice Scalia refers to ex abundanti cautela in comparably disparaging terms, as “a drafting imprecision venerable enough to have left its mark on legal Latin.”17 Yet at least one veteran legislator with nearly 50 years’ experience in state and federal legislatures dismisses arguments about redundancy as in general “wholly unpersuasive” and notes with satisfaction the frequency of repetition as a drafting technique.18 And in the Gluck and Bressman study, drafters identify practical and political reasons for often favoring insertion of redundancies rather than avoidance of superfluities.19 In pursuing a fuller understanding of the belt-and-suspenders approach, we come not to bury the rule against surplusage but to praise a new belt-and-suspenders canon when it is a rational interpretive strategy, keyed to legislative realities.

Accordingly, Part II makes an effort to show how widespread and self-conscious belt-and-suspenders practices are within statutory drafting, both at the federal and state levels. We identify two basic, and at times overlapping, reasons why such duplicative drafting occurs. One is out of an abundance of linguistic or purposive caution. Legislatures often seek to convey seriousness about their terms, to forestall doubts as to the meaning of crucial words or phrases, or to hammer home that a clarification in text either abrogates an incongruous court interpretation or does not abrogate a prior law. Earlier invocations of ex abundanti cautela acknowledged this reality, to be sure, but the desuetude of that canon has made judicial practice veer far from legislative practice. The other reason legislatures utilize belt-and-suspenders drafting is to build consensus among lawmakers. Sometimes using repetition facilitates forging compromises between bill drafters and any number of stakeholders who would prefer specific language in certain provisions, even if other provisions already cover the areas of stakeholder concern.

Part III explores how some courts have interfaced with this rather common modality of lawmaking within legislatures. Of course, we cannot canvass the innumerable instances in which courts adhere to their rule against superfluities and do not even consider the belt-and-suspenders counter-canon. Rather, we offer a range of case studies examining how courts come around to see the possibilities the belt-and-suspenders canon can suggest. This Part examines a few instances where courts have been willing to acknowledge and respect drafting repetitions but also explores cases where courts continue to dismiss as unpersuasive arguments invoking such repetitions.20

Part IV meditates on why there continues to be divergence between how legislatures and courts see redundant drafting practices. And then it aims to provide some guidance to courts for when they can stick to their guns with the old canon against superfluities—which is altogether too powerful in light of reasonable legislative preferences—and when they should embrace a legislature’s effort to fortify or bolster its statutes with belts and suspenders. This task is ultimately as urgent for “textualists” who want to get the text right as it is for “purposivists” or “intentionalists” who want to prioritize legislative design.

II.     Belt-and-Suspenders Legislative Drafting

Notwithstanding lamentations from the likes of Scalia and Garner that legislatures are lazy or loquacious—with a penchant for alliterative couplets or triplets—our study of legislative practices suggests that legislatures use “belt-and-suspenders” strategies to accomplish principled and rational objectives. Here we survey those strategies, examining the rather self-conscious legislative practice. We discuss these strategies below using two basic rubrics: caution-based drafting and consensus-based drafting.

A.     A Typology of Belt-and-Suspenders Drafting Techniques

1.     Caution

First, under the rubric of caution-based belt-and-suspenders drafting, legislators may be especially focused on textual exhaustiveness to make sure implementers get the message. The idea here is to reinforce the meaning of a term or provision. Some examples might include authorizing an agency to order that an employer engaged in wrongdoing “cease and desist” from its unlawful acts;21 creating a right of action for a minor who was “a victim of [certain sex crimes] and who suffer[ed] personal injury as a result;”22 or allowing a bankruptcy trustee to sell a debtor’s property “free and clear of any [other entity’s] interest in such property.”23 While these doublets might be understood as a single unit of analysis to convey only a single meaning, courts invoking the rule against superfluities do not always agree and sometimes seek to furnish independent meaning to each term.24 There may be a real irony here: Legislators might well be using very recognizable doublets—such as “cease and desist” and “aid and abet”—precisely because they worry about negative implications from trying to be too concise after a doublet takes hold.25

Apart from tightly formulated textual doublets, legislators sometimes draft redundantly in more expansive contexts within a single statute. Examples of this broader effort to double-cover may arise when Congress uses two separate sections of the same statute to make it a crime to “destroy[] . . . a record, document, or other object” with the intent of impeding an investigation;26 or when it defines “securities contracts” in several distinct yet overlapping ways as part of creating safe harbors in certain bankruptcy proceedings.27

Even more broadly, legislatures may opt for redundant drafting in relation to previously enacted statutes.28 Here, they may wish to be doubly sure they have not abrogated prior statutory protections, prohibitions, or requirements.29 Or they may add a new provision to assure that the text as augmented rejects a disturbing judicial interpretation of the previously-enacted text.30

These variations on redundant drafting share a motivational attribute: Congress or state legislatures are acting from an abundance of caution. Legislators repeat key terms or phrases in order to reinforce their meaning and importance—and in the process signal their emphasis to relevant public audiences. As part of exercising such drafting caution, legislators also seek to preclude any inclination by courts or agencies to take that meaning in a different interpretive direction. Indeed, Anglo-American common law jurisdictions have long recognized that legal formalism has had the effect of, in the words of Francis Lieber, “induc[ing] the law-makers to be, in their phraseology, as explicit and minute as possible.”31

2.     Consensus

Beyond the caution-based rationale, legislators also may adopt a belt-and-suspenders approach to drafting in order to attract colleagues who are fence-sitters or skeptics about the bill itself. A bill’s principal author or floor manager may believe that the text already covers some problem, but she will repeat the coverage once colleagues indicate that some kind of redundancy would increase their comfort level with the bill. These colleagues may come from within the legislative chamber, from a different cameral entity, from a different governmental branch, or even from other stakeholders in civil society.

When drafting redundantly to achieve consensus, a bill’s author or manager may opt to double-cover an issue within a single sentence or subsection, or between different sections, or across different statutes.32 In this respect, there can be overlap between consensus-building and caution-based approaches. Yet the two stem from analytically distinct rationales. Consensus-building is externally focused: It has political and policy dimensions that may involve communication across a spectrum of stakeholders. By contrast, a more purely caution-based approach tends to be internally focused: It is likely to stem from the perceptions or preferences of bill authors or legislative staff drafters, rather than from legislators’ dialogue in search of broader support.

Given this distinction, legislative record evidence articulating belt-and-suspenders drafting strategies is more apt to appear in the consensus-building context. Drafters who are doubly cautious on their own—without being pressed by colleagues—are less likely to announce themselves as pursuing a belt-and-suspenders approach. Most examples that follow from federal and state law involve some element of consensus-building discussion among legislators. Still, as noted above, the doubts expressed by fence-sitters or other colleagues may be attributable to textual caution as well as policy concerns. We attempt to convey our understanding of what underlies legislative preference for a belt-and-suspenders approach in each instance that follows.

B.     Belt-and-Suspenders in Congress

Our research has produced numerous instances of belt-and-suspenders drafting by Congress since the early 1990s. We offer below a set of examples that illustrate how common and deliberate this drafting technique is.

1.     When legislators worry about conflicts among statutory schemes

The Victims of Trafficking and Violence Protection Act became law in October 2000. The omnibus piece of legislation included provisions that had been part of an earlier-introduced bill addressing states’ enforcement of their alcohol laws under the Twenty-First Amendment.33 During the House floor debate on the earlier bill in August 1999, some legislators worried that injunctive relief authorized to prevent shipment of intoxicating liquor into a state in violation of state law might conflict with the then-recently enacted Internet Tax Freedom Act.34 In offering a bipartisan amendment to clarify that no such conflict existed between the House bill and the existing Internet statute,35 Rep. Cox expressed the “hope that this is a belt-and-suspenders operation.”36 He believed the bill language did not disturb prior law, but he saw the amendment as a re-emphasis to help bring at least one colleague on board. At the same time, Rep. Cox acknowledged that because certain states have been aggressive at taxing and regulating the internet, it was worth assuring “that no State confuses its power to tax or regulate alcoholic beverages with a new one found in this statute . . . to tax or regulate the Internet.”37 Once again, legislators’ embrace of a belt-and-suspenders rationale reflects an effort both to enhance support and to avoid an unintended alteration of a prior legal standard, focusing on a concern about a conflict among statutory schemes.

2.     When Congress needs to make sure longstanding policy does not get abrogated

A bipartisan bill in the House, introduced in 2017, was proposed to regulate the construction and maintenance of international border-crossing facilities for the import and export of oil, natural gas, and electricity.38 During the House floor debate, a bipartisan amendment was introduced to clarify bill language that some House members were concerned would abrogate the environmental review provisions of the National Environmental Policy Act (“NEPA”).39 Rep. Green, a primary sponsor of the bill and the amendment, referenced a CRS Report that had been introduced and discussed at a committee hearing on the bill; the report emphasized that nothing in the bill would take away the applicability of NEPA requirements. Green then added that because of the concerns expressed by his colleagues, “[w]e will make sure it is belts and suspenders and that [NEPA] is applied to these pipelines.”40 During floor discussion of the amendment specifying NEPA applicability, Green explained: “like I said earlier, it is belts and suspenders, but sometimes we need them to pass legislation.”41 The amendment and the bill passed the House,42 but the bill was not taken up in the Senate. The House floor discussion indicates a self-conscious use of belt-and-suspenders drafting for consensus-building purposes. At the same time, the addition of seemingly unnecessary language reinforcing NEPA applicability reflects a level of policy-related caution, to avoid an unintended abrogation of a pre-existing statutory requirement.

3.     Patent law and redundancy

As explained in a House Judiciary Committee Report accompanying a bill addressed to patentability in reexamination proceedings, a provision (ultimately enacted in 2002 as part of an Appropriations Act) made a small textual change to 35 U.S.C. §§ 303(a) and 312(a) in the nation’s patent law. It reiterated the basis for the U.S. Patent and Trademark Office (“PTO”) to determine whether a request for re-examination of a patent should be granted.43 The relevant prior text addressed to reexamination procedures gave the PTO Director discretion “[o]n his own initiative, and any time,” to determine and raise substantial new questions of patentability.44 A circuit court decision had rejected the Office’s discretion when a question arose involving “prior art that was before the PTO in an earlier examination.”45 This court decision apparently led to “gaming the system” by patent agents and lawyers who inserted hundreds of prior art references into their applications, knowing that PTO examiners were required to review the applications in compressed time frames.46

The bipartisan amendment added new text, which stated that “[t]he existence of a substantial new question of patentability is not precluded by the fact that a patent . . . was previously cited by or to the Office or considered by the Office.”47 As explained by the amendment’s principal sponsor:

This is what I call a belt-and-suspenders approach . . . . The goal is to allow reexamination of those cases where a genuine, substantial new question of patentability arises in light of prior art, which was reviewed by the Patent Office. At the same time, it leaves in place all of the protections for inventors . . . which exist under current law against frivolous or harassing conduct. While many believe the base text is satisfactory to meet that goal, I hope this removes any doubt . . . .48

One might regard this as simply an “override” amendment, given a circuit court holding that the committee report characterizes as “reach[ing] beyond the text of the Patent Act.”49 Yet, many members of the House Judiciary Committee apparently believed that reiterating the Office’s broad discretion through additional text made double-sure that PTO discretion would not be further questioned in the lower courts.

Another example of “belt-and-suspenders” drafting arose in connection with H.R. 760, introduced in 1993.50 Like a predecessor bill proposed in 1991,51 it aimed to expand patent protections for the U.S. biotech industry.52 A House hearing on the bill made clear that the drafters, acting in consultation with industry patent lawyers, adopted a self-consciously two-pronged approach.53 Title I expanded the definition of non-obviousness for certain biotechnological processes in order to supersede a recent decision by the Federal Circuit.54 Title II made biological starting material patents enforceable at the U.S. border.55 As explained by a leading industry witness, “[e]ither of the two prongs would solve the problem for the large majority of biotechnology inventions; together they would solve the entire problem.”56

Some subcommittee members wondered whether both provisions were necessary. Responding to a question from Representative Jack Reed, the Acting PTO Commissioner stated: “It is, in our view, something like belts and suspenders. If one amends section 103, one doesn’t need the latter approach and vice versa.”57 When Representative Reed followed up by observing, “[w]hat you seem to be saying is that you get a lot more protection in [this] bill than perhaps might be necessary. But what is wrong with more protection sometimes, I guess, would be my response?[,]” the PTO Acting Commissioner replied: “It is simply our view that you don’t need both, that is all.”58 In a later exchange with subcommittee Chair Hughes as to which title he would prefer if Congress determined the protections were duplicative, the vice president and associate general counsel of Amgen replied that while he would choose Title II, “we prefer both titles, [and] we feel there is nothing wrong with having belts and suspenders.”59

In the end, the bill as drafted did not pass the House. Nonetheless, this example and the previous patent law illustration are consistent with John Golden’s argument that patent law is an especially fertile area for the operation of redundant drafting and related conflicts with anti-redundancy presumptions.60 Insofar as the drafting of U.S. patent law is characterized by “relatively uncontroversial use of redundancy in institutional and process design,”61 it may be that traditional judicial suspicion towards redundant legal texts deserves special skepticism at least in the patent law context because the statutory drafters work in consultation with patent professionals who are themselves well-versed in the uses of belt-and-suspenders drafting techniques.

4.     When Congress internally disagrees about using belt-and-suspenders approaches

H.R. 2505, passed by the House in 2001 but not the Senate, was one of many efforts by House majorities over more than a decade to ban human cloning.62 The House Judiciary Committee Report includes a transcript of the committee markup during which Reps. Lofgren and Conyers proposed an amendment that nothing in the Act would prohibit stem cell research or therapies.63 Rep. Smith opposed the amendment, arguing that it was unnecessary because the bill’s language did not prohibit this type of research; Rep. Lofgren disagreed, citing support for her concerns from the National Institute of Health.64 At this point in the committee markup, Rep. Barney Frank expressed deep skepticism about “oppos[ing] an amendment solely because it is unnecessary.”65 Citing his decades of legislative experience, Rep. Frank observed that as politicians and lawyers, legislators are far from inherently opposed to repetition; indeed, they are “[t]he profession that has given the world ‘belt and suspenders’” among other redundant doublets. Rep. Frank added that bills are hardly great literature in which unnecessary language “spoils the rhythm of the prose,” and that arguments about redundancy should not inhibit debate on the merits of a proposal.66

Rep. Frank’s general position, that arguments against redundancy are a red herring, is less than fully persuasive. Superfluous language may contribute to confusion about textual meaning or even undermine language that seemed unambiguous prior to supplementation with a belt-and-suspenders amendment. For present purposes, though, Frank’s remarks usefully illustrate how legislators can and do have genuine bill-specific disagreements about (a) the value of belt-and-suspenders repetition when crafting statutory text; and (b) the possibility that arguments against redundancy will be used strategically to divert candid treatment of a proposal’s merits. Legislators are alert to the substance and nuance of these kinds of disagreements. But that substance and nuance may well be ignored or discounted by courts if they assume that legislatures are or should be committed to the rule against superfluities when drafting what might be considered redundant textual terms or provisions.

5.     When slack in the private regulatory process demands extra caution

The Senate version of what became the Consumer Product Safety Improvement Act in 2008 included a section on automatic garage door openers, mandating that all such openers include a protection device that did “not require contact with a person or object for the garage door to reverse.”67 During floor debate on the Senate bill, an amendment was offered to strike that section, on the ground that Congress should allow experts at the Underwriters Laboratories (“UL”) to continue to set the standard as they had done for years.68 Senator Pryor, bill author and floor manager, opposed the amendment.69 He explained that while UL had for many years promulgated a two-part safety standard (based on either contact or noncontact sensors), it had recently changed to only a contact-sensor approach.70 Pryor recognized there might be something “a little bit redundant” by insisting on both mechanisms, but he maintained that restoring the “belt-and suspenders” approach made sense where safety is concerned and in light of the fact that closing garage doors had caused accidental deaths.71 This single-product provision was ultimately removed in conference, although the Conference Report included language directing the CPSC to “expeditiously review, revise, and consider the adoption of standards” to prevent garage-door entrapment, “including contact and non-contact sensors.”72

This example suggests that members of Congress sometimes speak of “redundancy” and “belt-and-suspenders” drafting in a looser sense. The Senate here reverts to a prior two-pronged approach without noting that the specific statutory language of the two prongs is not itself redundant. Still, legislative drafters defended a belt-and-suspenders approach as an appropriately cautionary justification in this consumer safety setting, and as a way to push back against slack in the private regulatory process. The strategy indicates, once again, that legislators are prepared to invoke redundancy as a drafting asset in varied settings—although legislators do often self-consciously back away from belt-and-suspenders drafting when they feel it is counterproductive or otherwise contraindicated in a policy area.

* * *

What emerges from these federal illustrations—and they are far from exhaustive73—is that members of Congress are comfortable discussing and debating the use of belt-and-suspenders approaches to help justify various drafting choices. As we noted earlier, it is unsurprising that explicit references to belt-and-suspenders tend to arise in the context of efforts to generate consensus or solidify support. But sometimes even the “abundance of caution” justification is transparent in the legislative record. Ultimately, these references reflect a commitment to cautious drafting and an endorsement of repetition in order to reinforce textual meaning or policy preferences and to reassure colleagues. In short, members of Congress invoke belt-and-suspenders drafting strategies with some frequency, based on an understanding that the approach can be valuable. Even when it is controversial, it is treated as a legitimate strategy to promote caution and consensus.

C.     Belt-and-Suspenders in State Legislatures

Below we offer examples from two states—Connecticut and Texas—reflecting parallel themes to what we find in the congressional context. One of the two states (Texas) has codified the rule against surplusage while the other (Connecticut) has not.74 These examples reinforce our conclusions regarding the ways legislatures think about belt-and-suspenders drafting techniques.

1.     When an amendment is needed for caution and consensus

A 2013 Connecticut law specifies processes to assure that employees and former employees can have access to their personnel files upon request.75 The text reported to the floor by the House Labor Committee required the employer to “permit [a] former employee to inspect” her personnel file within ten business days of receiving such a request, not specifying whether the inspection would be in person.76 During debate in the House, a “friendly amendment” was offered stating that if there was no agreement on a location to conduct the inspection, the employer would satisfy the requirement by mailing a copy of the former employee’s personnel file to the former employee within ten business days of receiving the request.77 Rep. Tercyak, the Labor Committee Chair and floor manager, observed that “while some of us think that this amendment is sort of like wearing suspenders with a belt, what the heck, we’re the Labor Committee and we’re working hard to bring people together.”78 The amendment was approved and the amended bill became law later that year.79

Before Rep. Smith’s amendment, the bill did not indicate that an inspection had to be in person. And because this section dealt with former employees who might no longer reside in the vicinity, one might have inferred that in-person inspections were not routinely anticipated. This presumably is why Rep. Tercyak invoked the belt-and-suspenders analogy. That said, his decision to embrace the amendment reflects an effort to build or maintain consensus—an effort that was rewarded.

2.     When a legislature needs an exclamation point!

In August 2016, when the state legislature was not in session, “Connecticut was awarded a federal grant to launch a pilot mileage tax program so long as the state invested $300,000.”80 The idea of a mileage tax study had been floated in 2015 at a meeting of the governor’s transportation finance panel (as a way to help pay for the governor’s $100 billion transportation plan), but taxpayers and the trucking industry expressed strong resistance and the Senate Transportation Committee concluded that no such study was needed.81

There was an outcry following awareness of the 2016 federal grant, with legislators expressing anger about what they deemed “bureaucratic overreach” to study a tax that “hadn’t been debated or even authorized by the legislature.”82 The legislature then enacted a law making it clear that the state Department of Transportation was not authorized to spend any money directly or indirectly on a study, conference, or any other activity related to a potential mileage tax.83 Speaking in support of the proposed law, Sen. Looney described the bill as “a reemphasis, a doubling of emphasis—it’s in effect, adding suspenders to a belt to point out that in fact, we are as a body, vehemently opposed to the concept of a vehicle mileage tax.”84

This example does not reflect classic belt-and-suspenders drafting, inasmuch as there was not an initial statutory text, just a Senate Committee conclusion of rejection. Still, invoking the belt-and-suspenders image to support a doubling down against agency overreach is a reminder of how comfortable legislatures have become with the metaphor as well as the drafting practice.

3.     When state legislatures worry about impairments of core state values

A 2017 Texas law, authorizing the continuation and functions of the state bar through August 2029, includes a provision added as an amendment during floor debate, directing the Texas Supreme Court to “ensure that no rule [governing admission to the practice of law] violates Chapter 110, Civil Practice and Remedies Code.”85 That chapter in turn prohibits government agencies from substantially burdening a person’s free exercise of religion.86 Amendment supporters sought to negate the possible effect of an ABA model rule, which they viewed as unduly expanding the definition of professional misconduct involving religious harassment or discrimination.87

During House floor debate on an initial version of the amendment, Rep. Leach, the amendment sponsor, maintained that it was needed to ensure that the Texas Supreme Court followed First Amendment precedents rather than an ABA model rule.88 When a colleague responded that he hoped the state supreme court would always follow the Constitution, Leach replied: “I would hope the same thing, Representative Turner. I’m glad we agree on that, but this amendment is belts and suspenders for any future supreme court that might be tempted to follow the American Bar Association rules which are clearly unconstitutional.”89 Although one suspects a degree of strategic posturing on Leach’s part,90 the exchange indicates that legislators are prepared to invoke belt-and-suspenders strategies in an effort to preempt future judicial interpretations.

4.     When a state legislature reveals some of its anxieties about its work product

A 2011 Texas bill added a set of informed consent requirements in order to make it harder to obtain an abortion in the state.91 A relatively standard severability provision amendment was proposed and debated on the House floor.92 Rep. Miller opined that while his amendment might not be necessary, its inclusion was an important backstop against any oversights in the bill. Rep. Farrar suggested that Miller’s express insistence on the use of belt-and-suspenders drafting—“Actually, I really don’t have any concerns. But it’s just kind of like putting on your belt and suspenders, too. I want to cover all aspects.”93—revealed that Miller was less than fully confident in the bill as drafted. Nonetheless, the supposedly unnecessary severability clause was included to help preserve the rest of the bill and make its intent clear. Ultimately, the clause became part of the final bill, which was signed by the Governor in 2011.

* * *

These examples from state legislatures, like the congressional ones before them, highlight that legislatures self-consciously embrace belt-and-suspenders drafting—and utilize it to build consensus and to promote caution and clarity. Sometimes the relevant audience for the belt-and-suspenders approach is the judiciary; sometimes it is an agency; and sometimes it is there for a constituent or other stakeholder. In all these state cases, however, legislatures have rationales for what they are doing; the redundancies do not reflect laziness or an effort to design a neat turn of phrase with a doublet or triplet. Part III below explores several cases in which judges think about how to come to terms with the reality that legislatures often employ belt-and-suspenders drafting approaches.

III.     Belt-and-Suspenders in the Courts

Given the range of legislative examples and motivations we explored in Part II, belt-and-suspendering does seem like a background norm and convention among statutory drafters. But appreciating this legislative reality better also requires integrating it into the pantheon of legitimate interpretive canons.94 A belt-and-suspenders canon in the courts would be salutary because of the practice’s clear functional justifications as part of legislative drafting,95 the interpretative approach’s ample historical pedigree,96 and owing to frequent invocation by justices and lower court judges. Indeed, these rationales qualify a belt-and-suspenders canon for membership even under rigorous entry criteria.97 Moreover, its routine competitor canon, a presumption of non-redundancy, lacks the bona fides of mirroring a drafting reality that we have demonstrated above exists across a spectrum of settings. Although it is appropriate to recognize that the anti-surplusage canon reflects a judicial elevation of the value of clarity, this must be integrated with the legislative values of caution and consensus.98 And those who keep the faith and spread the gospel of the canons have recognized a “principle of interrelating canons,” that “no canon of interpretation is absolute. Each may be overcome by the strength of differing principles that point in other directions.”99

Because courts are so familiar with the rule against superfluities, they have been slower to respond to how legislatures use redundancies when producing their statutes. Yet judges are sometimes shaken out of their preference for (and effort to enforce) parsimony by legislators. In such cases, they may be willing to credit belt-and-suspenders drafting techniques, or at least constructively consider such possibilities. Perhaps we can draw lessons from some of the times courts have engaged these issues in the federal and state courts.

Below, we focus first on some relatively uncontroversial cases. We then fan out to harder cases in which courts appear more conflicted. In doing so, we consider how the more difficult cases might help courts to engage this common drafting technique in the future. The easy cases may arise in the form of endorsing the relevance of belt-and-suspenders drafting and the irrelevance of the rule against superfluities.100 Alternatively, the easy cases can be reasonable rejections of belt-and-suspenders arguments in favor of the rule against superfluities.101 But sometimes the cases are just hard.102 After discussing examples of all three kinds of cases, we offer observations as to how the presumption against surplusage might be reshaped.103 We do so in light of our findings in Part II that a belt-and-suspenders canon would reflect a common and self-conscious drafting technique, drawing judicial practice closer to legislative practice.

A.     Rejecting Anti-Redundancy Norms in Favor of Recognizing Belt-and-Suspenders Drafting

For a clear example in the Supreme Court, consider the 9–0 decision by Justice Scalia in Freeman v. Quicken Loans.104 There, the Court was unwilling to try to give distinct meanings to the words “portion, split, or percentage” in the Real Estate Settlement Procedures Act (“RESPA”).105 RESPA provides that “[n]o person shall give and no person shall accept any portion, split, or percentage of any charge made or received for the rendering of a real estate settlement service . . . other than for services actually performed.”106 But rather than employing “the “canon against surplusage,” the Court decided that the legislature was using a “not uncommon sort of lawyerly iteration,” finding it “impossible to imagine a ‘portion’ . . . or a ‘split’ that is not also a ‘percentage.’”107 The Court does (regrettably from our standpoint) call this kind of reiteration “perhaps regrettable.”108 But notwithstanding its judgment that it would prefer less repetitive drafting, it seems to acknowledge that the reality of legislative drafting is not wholly consistent with the anti-surplusage assumption so often enforced by courts.109

Another recent example occurred in the Court of Appeals for the D.C. Circuit in United States v. Bronstein.110 The case stemmed from a set of pranksters who sought to disrupt U.S. Supreme Court proceedings, after having been warned to stay quiet by Chief Justice Roberts. They were arrested under a statute which provides that “[i]t is unlawful to discharge a firearm, firework or explosive, set fire to a combustible, make a harangue or oration, or utter loud, threatening, or abusive language in the Supreme Court Building or grounds.”111 Defendant Bronstein was successful in the lower court in convincing the judge that the words “harangue” and “oration” were unconstitutionally vague.112

In rejecting the lower court’s view on appeal, the D.C. Circuit found the words to have “a settled meaning around public speeches[,]”113 and utilized other canons of interpretation—noscitur a sociis and the rule that a statute’s title may provide clues—to reinforce the meaning of those words from other associated words in the statute.114 Although the lower court had “viewed the convergence of ‘harangue’ and ‘oration’ on a single meaning as indicative of their respective vagueness[,]” the appellate court ultimately discounted the potential redundancy in the statute.115 Quoting Scalia and Garner, the court held that “[s]ometimes drafters do repeat themselves and do include words that add nothing of substance, either out of a flawed sense of style or to engage in the ill-conceived but lamentably common belt-and-suspenders approach.”116 Although unnecessarily adopting Scalia’s (and Garner’s) frustration about the recalcitrance of belt-and-suspenders drafting in legislatures,117 the court here recognizes and credits the technique, subordinating what it calls the “surplusage canon.”118 The court concludes: “When a statute’s text, context, and history all converge on certain terms possessing a settled legal meaning, the Court should effectuate it. The alternative—following a presumption of legislative precision [and parsimony] over the Constitution’s precipice—does not vindicate substance. It privileges theory.”119Bronstein’s clarity has also productively led other courts to be more cautious about rote invocations of the anti-redundancy canon.120

It is not only when courts spot synonym strings that they can be willing to acknowledge the use of belt-and-suspenders drafting, although such strings do seem to be an archetype for suppressing the force of anti-redundancy norms.121 When courts spot overlapping subparts of bigger statutes, they can be willing to enforce a natural reading of a statute over trying to stay true to a pure rule against superfluities.122 They also can be willing to see the import of belt-and-suspenders drafting even when it would conflict with a rigid commitment to giving effect to all meaningful variations of statutory language.123

B.     Rejecting Belt-and-Suspenders Arguments

These cases notwithstanding, there are also relatively easy cases where courts properly identify that belt-and-suspenders drafting is not a likely explanation for statutory language. This reinforces that courts can responsibly utilize the canon—and that the rule against superfluities wouldn’t need to be wholesale rejected to admit belt-and-suspenders reasoning into mainstream statutory interpretation.

In Rajala v. Gardner, for example, a Tenth Circuit panel had to decide whether an allegedly fraudulently transferred asset should be considered part of a bankruptcy estate before it had been recovered by the trustee.124 If so, the asset would be covered by the automatic stay of “any act to obtain possession of property of the estate” triggered by a filing of a bankruptcy petition under § 362(a)(3) of the Bankruptcy Code.125 Section 541 defines property of the estate to include, in (a)(1), “all legal or equitable interests of the debtor in property as of the commencement of the case,” and, in (a)(3), “[a]ny interest in property that the trustee recovers under section . . . 550” of the Code.126 In turn, § 550 enables the trustee to recover for the estate’s property “to the extent that a transfer is avoided under” provisions that give the trustee authority to avoid fraudulent transfers in § 548.127 The Rajala court had to decide whether allegedly fraudulently transferred property should count as “property of the estate” under § 541(a) before that property is recovered under § 550(a) (thereby subject to the stay imposed by § 362(a)).128

The court relied on the “cardinal principle of statutory construction that . . . if it can be prevented, no clause, sentence, or word shall be superfluous, void, or insignificant.”129 The court deemed that “interpreting § 541(a)(1) to include fraudulently transferred property would render § 541(a)(3) meaningless with respect to property recovered in a fraudulent transfer action.”130 Although the trustee had argued “that § 541(a)(3) is ‘a belt and suspenders,’ designed to ensure that assets will be available to satisfy creditor interests[,]” the court didn’t buy it and stuck with plain meaning.131 The trustee emphasized that the belt-and-suspenders drafting might have “act[ed] as a deterrent to fraudulent transfers and furthers the bankruptcy objectives of asset preservation and equitable distribution.”132 The court recognized the existence of some legislative history supporting this position, but found contrary evidence of intent in the statute’s structure.133 In the end, the court pointed to many other mechanisms for safeguarding debtor assets and reinforced a commitment in the case to anti-redundancy norms.134 Although this issue of interpretation has split the circuits,135 only the Rajala court considered and reasonably rejected the belt-and-suspenders argument.

Another example with which anyone who studies statutory interpretation would be familiar is Babbitt v. Sweet Home.136 At issue there was whether the Secretary of Interior properly interpreted the word “harm,” as used in the statutory definition of the word “take” in the Endangered Species Act (outlawing takings), to include certain kinds of habitat modifications that had the effect of killing or injuring wildlife.137 The statutory definition of “take” encompassed a list of proscribed forms of taking: one cannot “harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect” endangered species. All of the proscriptions—other than “harm”—for the most part seem like they target direct injuries to endangered species.138 Thus, the Court had to decide whether the definition’s inclusion of the word “harm” added anything to that list or instead should be deemed to overlap substantially with the rest of its statutory neighbors.

The Court could have embraced a belt-and-suspenders argument that would have veered close to a noscitur a sociis analysis, limiting “harm” to essentially the same kind of conduct otherwise covered by statutory neighbors in the definitional section. Justice Scalia urged this position in a vehement dissent,139 following the lower court decision for the D.C. Circuit.140 Instead, the Court properly preferred an anti-surplusage argument,141 giving “harm” new content that vindicated the Secretary of Interior’s interpretation.142 Justice Stevens’ majority opinion, however, didn’t fetishize parsimony in its embrace of the rule against superfluities. Rather, the Court relied heavily on drafting history and other signals of textual meaning.143 Although litigants and judges did not raise the belt-and-suspenders canon in quite those terms, the Court convincingly rejected the gist of that kind of reasoning in this context.

C.     Hard Cases

And then there are harder cases, where courts are internally split about the relative power of an anti-redundancy norm as compared with the commonsense consideration that legislatures draft with belt-and-suspenders techniques all the time. Consider in this regard State v. Nelson from the Supreme Court of Minnesota.144 There, the court had to consider what to do about a state statute that criminalized a person’s failure to provide “care and support” to a spouse or child when legally required.145 Nelson, convicted under the statute, claimed that the state had not met its burden to prove that he failed to “care” for his children, though he conceded that he was not paying required “support.”146 Somewhat surprisingly, the Supreme Court agreed with Nelson that the “care-and-support” statute required the state to show both that Nelson failed to provide care and that he failed to provide financial support.147

Justice Stras (who has since been appointed by President Trump to the Eighth Circuit) found that the statute is ultimately ambiguous: The statute either could be criminalizing the failure to pay support and is indifferent to whether a nonpayer is also failing in nonmonetary care obligations; or it could be imposing criminal sanction only to offenders who are delinquent on both fronts.148 Justice Stras’s majority deployed many canons, trying to adjudicate between these two readings.149 But it settled on an anti-redundancy preference “requiring [the court] to give meaning to every word and phrase in a statute” to generate the ambiguity,150 and then used the “rule of lenity requir[ing] [the court] to resolve the ambiguity in the care-and-support statute in favor of the criminal defendant.”151 It is, perhaps, notable that Minnesota is one of ten states that have codified its rule against surplusage152—and Justice Stras cited that state statute in support of his anti-redundancy reading.153

The dissent offered a plausible counter-reading, relying inter alia on what we would call the belt-and-suspenders canon.154 Citing Freeman v. Quicken Loans155 with which we began in Part III, the dissent argued that “reliance on the presumption against surplusage is misplaced in this case.”156 Instead, the dissent claimed that “care and support” is a “doublet,” referencing a work by Bryan Garner that “list[s] ‘over 100 examples of doublets.’”157

Yet Garner’s list doesn’t actually include this doublet, which raises the concern that the dissent is calling this language a doublet without really establishing it is one. Whether it is such a doublet is, after all, the very question at issue. The dissent’s further reliance on “the fact that Minn. Stat. § 609.375, subd. 1, contains a second doublet in the phrase ‘knowingly omits and fails’”158 seems like a non sequitur; the dissent offers no explanation for how one doublet breeds another (though the general point that legislatures do not use perfectly precise non-repetitive language is virtually unassailable).

We have some sympathy for the claim that “care and support” is a belt-and-suspenders formulation. Indeed, it seems to us likely that the dissent was correct, notwithstanding a codification presumptively favoring the anti-surplusage canon within the state’s law. And the dissent’s view was in part vindicated by the Minnesota legislature’s swift override of Justice Stras’s opinion very soon after the decision.159 But Nelson remains a hard case because Stras is persuasive that the dissent uses a belt-and-suspenders argument to minimize the complexity of deciding the very thing at issue—whether support and care are two things or one thing. This fact gives the lie to Scalia and Garner’s certainty that “when a drafter has engaged in the retrograde practice of stringing out synonyms and near-synonyms (e.g., transfer, assign, convey, alienate, or set over), the bad habit is so easily detectible that the canon [of the rule against superfluities] can be appropriately discounted: Alienate will not be held to mean something wholly distinct from transfer, convey, and assign, etc.”160

Identifying doublets and the right units of analysis in statutes is not the only challenge that makes for hard cases. There are also difficult arguments about how belt-and-suspenders reasoning works when comparing statutes that have seemingly meaningful variation. For example, consider a debate that arose on a Tenth Circuit panel between then-Judge Gorsuch and Judge Lucero in United States v. Smith.161 At issue was a mandatory minimum sentencing statute—and whether a sentencing judge is allowed to consider that mandatory sentence for using a gun during a crime of violence when devising an appropriate sentence for the underlying crime of violence.162 In Smith’s case, he was convicted of two counts of robbery163 and two counts associated with his use of “a gun ‘during and in relation to’ those ‘crime[s] of violence.’”164 At sentencing, the district court judge, after imposing a 35-year mandatory sentence on the gun charge, was encouraged by the government to disregard this lengthy gun sentence as it turned to weigh the appropriate additional prison term for the robberies themselves.165

Gorsuch’s majority opinion remanded back to the sentencing court because he found the government’s argument unpersuasive. Utilizing a wide range of interpretive canons (“specific provisions control general provisions”! “no elephants in mouseholes”! “watchdogs did not bark in the night!” “parsimony, not surplusage!” “meaningful variation!”),166 the court read 18 U.S.C. § 924(c) to require the sentencing court to impose consecutive rather than concurrent sentences for the underlying crime of violence—but found no statutory command that would require the sentencing court to close its eyes to the gun calculation when turning to the use of traditional sentencing discretion for the underlying crime of violence.

To reinforce its conclusion, the court looked carefully at a “statutory cousin” in 18 U.S.C. § 1028A, which criminalizes identity theft “during and in relation to” certain other crimes.167 This cousin also requires that the penalties it imposes must be “in addition to the punishment provided for” the underlying crime and that the penalties be consecutive rather than concurrent.168 But the cousin also includes 18 U.S.C. § 1028A(b)(3), a subsection that is explicit in its requirement that “in determining any term of imprisonment to be imposed for the felony during which the [identity theft occurred], a court shall not in any way reduce the term to be imposed for such crime so as to compensate for . . . any separate term of imprisonment imposed or to be imposed for a violation of this section.”169 Gorsuch essentially argued that Congress knows precisely how to write a statute to prohibit sentence reductions, which is what the government claimed to be the best reading of § 924(c). Yet Congress chose not to do it—and construing
§ 924(c) to accomplish this objective would, in effect, “render § 1028A(b)(3) superfluous.”170

It is from the argument about the “statutory cousin”—and particularly the explicitness of § 1028A(b)(3)—that Judge Lucero dissented, in part:

Congress may include technically unnecessary language out of an abundance of caution under the canon ex abundanti cautela. Indeed, the legislative history of § 1028A indicates that Congress directed the mandatory minimum to be imposed “in addition to any term of imprisonment for the underlying offense,” and included subsection (b)(3) “to ensure the intent of th[e] legislation is carried out.” That Congress chose a belt-and-suspenders approach in one statute does not render suspenders alone insufficient [in another].”171

By drawing upon what we are calling the belt-and-suspenders canon (and its historical predecessor, ex abundanti cautela), Judge Lucero tried to impress upon the court that the redundancy common to statutory drafting (as in
§ 1028A) should not be used to undermine more parsimonious but also effective language (as in § 924(c)) that accomplishes the same ends: in this instance, disabling courts from considering the mandatory minimum sentences when turning to impose a penalty for the underlying crime.172

On the one hand, Lucero is surely right that the reality of drafting practices needs to inform how courts read statutes, especially when they are trying to divine meaning from “statutory cousins” and other parts of the legislative code. At the same time, the rule that encourages courts to look for meaningful variation between two statutory schemes—especially when they are as related as they are here—also seems like a sound and reasonable approach to divining meaning. And yet, doubling down on a “cardinal principle” “to give effect . . . to every clause and word of a statute,” as Gorsuch does in Smith,173 seems remarkably fetishistic. In the end, knowing when belt-and-suspenders reasoning should control over the rule of meaningful variation may well be as much art as science. But it seems clear that turning a blind eye to the routine practice of belt-and-suspenders drafting in the legislature is ill-advised.

Our final challenging setting involves a provision of the National Voting Rights Act (“NVRA”) that requires states to “accept and use” a federal voter registration form developed by the Election Assistance Commission, although states are permitted to use their own form as well.174 In Gonzalez v. Arizona, the Ninth Circuit, sitting en banc, had to decide whether this “accept and use” provision meant that states could not add their own requirements to the federal registration form.175 The court held that the NVRA federal form set a standard from which states could not depart.176

Of interest to us here is then-Chief Judge Kozinski’s concurring opinion, which relies on legislative history to resolve what he viewed as a textual ambiguity: whether “accept and use” has an exclusive rather than inclusive meaning.177 If, contrary to the majority, the national form could be construed as one component in the state’s requirements to register, the federal form could then be seen as a “belt” to which the states could add regulatory “suspenders.”178 Some early legislative history cut against the exclusive reading. Senator Ford (the bill sponsor), responding to a proposed amendment from Senator Simpson that would have explicitly allowed states to choose to add documentary requirements, described the amendment as “basically . . . redundant” because nothing in the bill would preclude such state action.179 Senator Ford did not object to what he regarded as a superfluous amendment, and it was unanimously agreed to in the Senate.180

The House version of the bill had no such provision, however, and the Conference Committee adopted the House version. The Committee report explained that the Senate amendment was “not necessary or consistent with the purposes of this Act.”181 The conferees expressed their concern that the amendment “could be interpreted by [the] States to permit registration requirements that could effectively eliminate, or seriously interfere with, the mail registration program of the Act [and] could also adversely affect the administration of the other [federal] registration programs as well.”182 Judge Kozinski concluded that “[t]he conferees thus rejected the Simpson amendment, not because they thought it [belt-and-suspenders] superfluous (as did Senator Ford) but because the inclusive meaning of ‘accept and use’ was inconsistent with their vision of how the Act should operate.”183

* * *

Part III has shown that the belt-and-suspenders canon is in modest use in courts—but our analyses here focus on how courts can more effectively apply this interpretative tool. To be sure, Part II finds more widespread use of
belt-and-suspenders drafting than seems to be reflected in judicial understandings. This begs for some discussion about why there is divergence between what legislatures are doing and judicial understanding of that activity, and also invites some observations and recommendations for best practices with this interpretive canon going forward.

IV.     Analyzing Belt-and-Suspenders

Here we bring the strands of our investigations together to offer some ruminations about both the legislative drafting technique and the inchoate canon in the courts, where judges are starting to acknowledge legislative drafting realities. First, we reflect on what explains the continuing and stubborn divergence between the centrality of belt-and-suspenders drafting techniques in the legislature and its marginality as a canon of interpretation in the judiciary (Section IV.A). We then we turn to what we imagine to be a better future for the canon (Section IV.B).

A.     Continuing Legislative-Judicial Divergence

There seems to be a basic reason that explains the continuing divergence between courts and legislatures in this space: Legislatures and courts have different primary audiences. Legislatures see themselves as working together to convince one another and stakeholders to embrace important policy objectives (what we have called in Part II a “consensus-based” technique),184 and also talking to judges and agencies to make “double sure” their policy objectives are met (what we have called in Part II a “caution-based” technique).185 By contrast, judges view themselves as rule-of-law promoters in the service of lawyers and ordinary folk. This role calls upon courts to interpret enacted text to be as parsimonious and clear as possible. It is thus not especially surprising that a culture clash of sorts has created a tension between legislatures and courts.

In striving for clarity, however reasonable in the abstract, courts too often see themselves as involved in the process of chastening and chastising their statutory bosses to write in more meticulously clear terms.186 Consider Scalia (with Garner, again) here: “Statutes should be carefully drafted, and encouraging courts to ignore sloppily inserted words results in legislative freeriding and increasingly slipshod drafting. Nothing should be included in a legal instrument ‘for no good reason at all.’”187 And they double-down: “if the legislators themselves are not mindful of ferreting out words and phrases that contribute nothing to meaning, they ought to hire eagle-eyed editors who are.”188 Although Scalia and Garner obviously can’t speak for all judges, we suspect they are giving expression to why judges remain more hostile to belt-and-suspenders legislative techniques than we believe they should be, given its justifiable prominence in legislative practice. Indeed, many of the cases we examined that consider the canon ultimately refer back to Scalia and Garner, quoting their language that finds this drafting technique “flawed,”189 “ill-conceived,”190 “lamentabl[e],”191 “retrograde,”192 a “bad habit.”193 Yet if the legislature has rationales for this behavior that do not stem from sloppiness, it is hard to see why we should promote judicial supremacy in this domain, allowing the judiciary to lecture the legislature about how the latter should pursue—or compromise over—policy objectives as part of a complex and pressured lawmaking process. Indeed, belt-and-suspenders drafting can itself also be directly responsive to judicial efforts to demand clarity—repetition is one routine modality of emphasis—so it is ironic that legislative responsiveness to signals demanding clarity can too easily be ignored if the courts continue to press their anti-redundancy norm against the belt-and-suspenders reality of legislative practice.

As we developed in Part II, legislatures are rather self-conscious and thoughtful about their belt-and-suspendering. They use this drafting technique as part of their communicative enterprise, as they deliberate internally and communicate externally. When judges seek to take this strategy away through a baroque effort to demand parsimony, they are derogating from the project of interpretation, which should be, centrally, making a good faith effort to implement what the democratically elected statutory bosses enact. They are also being insensitive to the hard work it takes to set democratic policy in motion. Even largely textualist jurists like Kavanaugh and Kozinski have warmed to the idea that judges need to be more respectful of how legislatures actually function and do their work.194 We hope that in exposing some of the details and nuances associated with the belt-and-suspenders drafting technique, there will be broader respect for the application of a belt-and-suspenders canon going forward.

B.     The Future for a Belt-and-Suspenders Canon

The future for the canon, we suspect, will be a net softening of the rule against superfluities. That norm against surplusage is too well-entrenched and powerful to disappear completely, though we’d probably encourage the ten states that have enacted some version of the anti-redundancy norm either to repeal it or to add a belt-and-suspenders proviso in their codes. Even if courts aren’t always paying attention to legislative direction (as the Minnesota majority did in Nelson above),195 it would be best to have codified canons match up better with drafting practices. But the more courts get comfortable with legislative realities associated with belt-and-suspenders drafting, the more we can reasonably expect a mitigation of the too-often-deployed anti-redundancy norm, which is effectively a product of judicial invention rather than of legislative preference-estimation. Our exploration of the comfort level courts have shown with accepting and rejecting a belt-and-suspenders canon reinforces the conclusion that it can be applied in responsible ways. Yet having wrestled with some harder cases in Part III, we offer a few specific observations toward more principled use in cases to come.

First, the belt-and-suspenders canon needs to be foregrounded, not treated as a least-best alternative. As we have suggested, belt-and-suspendering isn’t, in the average case, sloppy, lazy, or regrettable. Accordingly, judges need to stop talking and thinking that way; instead they should consider legislative dynamics of caution and consensus before they resort to anti-surplusage norms. Courts cannot demand clarity out of one side of their proverbial mouths and then stick their heads in the sand about the ways legislatures work to get clear.

Second, one valuable approach to differentiating between belt-and-suspenders on the one hand and anti-surplusage on the other is to look for clues in legislative history. Not every judge is comfortable with reverting to legislative history, of course. But it is worth noting that even judges who prefer textualism might, in the kinds of cases for which canons become relevant, appreciate that ambiguities should trigger some resort to drafting history that can help clarify meaning.196 For the textualists who have exclusionary instincts about legislative history, however, honesty about the legislative process probably requires picking belt-and-suspenders readings in a larger proportion of cases over anti-surplusage readings, assuming there is little else to help the judge choose a reading. Recognizing a belt-and-suspenders canon is not necessarily preferring intent to text. Rather, belt-and-suspenders readings are efforts to discover the communicative content of the statute. If anything, anti-surplusage readings are more likely to be driven by judicial norms rather than the meaning of the text.

To be sure, the belt-and-suspenders canon—like any canon—is not always going to be dispositive in producing a clear reading. As we highlighted in our discussion of Nelson in Part III,197 doublets and synonym strings are some evidence of belting-and-suspendering. But it risks arguing by fiat to identify any two closely related words to be an instance of belt-and-suspenders. Far from always being “so easily detectible” as Scalia and Garner suggest,198 many cases involve good arguments on both sides. For the textualist set like Stras in Nelson and Gorsuch in Smith, the couplet of anti-redundancy and belt-and-suspenders could lead to greater deployment of the rule of lenity to resolve ambiguities that otherwise dueling canons cannot clarify, as indeed occurred in those two cases. This is not obviously a faulty ranking of the various canons.199 But for the textualists willing to do a little bit of peeking at legislative history, the record will sometimes help the judge choose between anti-surplusage and the belt-and-suspenders canon.

It is high time, however, that we stop wishing for perfectly parsimonious drafting and do a better job trying to uncover whether statutory provisions are there to be caution-inducing or consensus-generating on the one hand, as opposed to an effort to capture different dimensions of a prohibition or command without surplusage. Now that we have explored much of the logic generating belt-and-suspenders drafting in legislatures, we hope courts will extend their use of the canon, and will provide deeper arguments for when its application should be respected and when it should be suppressed in favor of another canon.

Finally, our discussion of Smith makes clear that doublets and synonym strings are not going to be the only area of complexity for the belt-and-suspenders canon.200 The study of that case highlights the care that must be taken with the canon’s application when courts undertake comparisons of statutes throughout the code, looking for hints about meaning by examining similar statutes to the one under investigation, to see how the legislature took to drafting in related contexts. While being mindful of the possibility for belt-and-suspenders design is essential, it is also critical that courts pursue context and drafting history here, too, to deploy the belt-and-suspenders canon properly.

V.     Conclusion

As a benchmark for biblical exegesis under conditions of divine authorship, it might make sense to think that every word in God’s law has a distinctive meaning and that even a prolix legal code by an omniscient lawgiver would be fashioned to have little surplusage.201 But as applied to human law, drafted by dozens of people with their hands in the sausage-making, it remains difficult to pull from the rules of religious exegesis to earthly correlates. Although this isn’t the place to argue for displacing the secular rule against superfluities completely, we believe we have exposed a powerful counter-canon that has lots of support among those who actually do the work of drafting and approving statutes. With more attention to the belt-and-suspenders drafting techniques—and their justifications—courts engaged in statutory interpretation can hope to do better in their deployment of what is at present a too-powerful norm against redundancy.

  1. [1]. Yates v. United States, 135 S. Ct. 1074, 1096 (2015) (Kagan, J., dissenting) (“The presence of both § 1519 and § 1512(c)(1) in the final Act may have reflected belt-and-suspenders caution.”).

  2. [2]. Hively v. Ivy Tech Cmty. Coll. of Ind., 853 F.3d 339, 344 (7th Cir. 2017) (en banc) (“Congress may choose a belt-and-suspenders approach to promote its policy objectives.” (quoting McEvoy v. IEI Barge Servs., Inc., 622 F.3d 671, 677 (7th Cir. 2010))).

  3. [3]. See also Rimini St., Inc. v. Oracle USA, Inc., 139 S. Ct. 873, 881 (2019) (“Redundancy is not a silver bullet... . Sometimes the better overall reading of the statute contains some redundancy.”); King v. Burwell, 135 S. Ct. 2480, 2498 (2015) (Scalia, J., dissenting) (“Lawmakers sometimes repeat themselves—whether out of a desire to add emphasis, a sense of belt-and-suspenders caution, or a lawyerly penchant for doublets (aid and abet, cease and desist, null and void).”).

  4. [4]. See Circuit City Stores, Inc. v. Adams, 532 U.S. 105, 140 (2001) (Souter, J., dissenting) (“[T]he explanation for the catchall is not ejusdem generis; instead, the explanation for the specifics is ex abundanti cautela, abundance of caution.” (citing Fort Stewart Sch. v. FLRA, 495 U.S. 641, 646 (1990))); Seven-Sky v. Holder, 661 F.3d 1, 37–38 (D.C. Cir. 2011) (Kavanaugh, J., dissenting) (“The lesson from the redundancy in these sections ... is not to read provisions out of the statute or contrary to their plain meaning, as the majority opinion would have us do. Rather, we should read the provisions according to their terms, recognizing that Congress often wants to make ‘double sure’—a technique so common that it has spawned its own Latin canon, ex abundanti cautela.”); Marx v. Gen. Revenue Corp., 668 F.3d 1174, 1183 (10th Cir. 2011) (“It may seem redundant, but if canons of construction are to be invoked, the appropriate one is that of ex abundanti cautela (abundance of caution), which teaches that Congress may on occasion repeat language in order to emphasize it.”); United States v. Bendtzen, 542 F.3d 722, 727 (9th Cir. 2008); United States v. Kaluza, 780 F.3d 647, 658 (5th Cir. 2015).

  5. [5]. The lineage of the ex abundanti cautela canon in the United States traces back to some of our most famous justices. See United States v. Wiltberger, 18 U.S. (5 Wheat.) 76, 115 (1820) (Marshall, C.J.); Brown v. United States, 12 U.S. (8 Cranch) 110, 150–51 (1814) (Story, J., dissenting). For other earlier uses, see Childress v. Emory, 21 U.S. (8 Wheat.) 642, 663 (1823); Bank of Hamilton v. Lessee of Dudley, 27 U.S. (2 Pet.) 492, 502 (1829); Manhattan Co. v. City of Ironwood, 74 F. 535, 540 (6th Cir. 1896) (Taft, J.); In re New Amsterdam Motor Co., 180 F. 943, 944 (S.D.N.Y. 1910); In re John Liddle Cut Stone Co., 242 F. 691, 693 (S.D.N.Y. 1916); In re Toole, 294 F. 975, 977 (S.D.N.Y. 1920); Comm’r v. Van Schaick, 83 F.2d 940, 942 (2d Cir. 1936); Founders General Corporation Corp. v. Hoey, 84 F.2d 976, 978 (2d Cir. 1936). We don’t know why so many of these cases from S.D.N.Y. and the Second Circuit are penned by Augustus Hand. And he rejects the reasoning about as often as he embraces it.

              Crediting a legislature’s efforts in statutory drafting to “make assurance doubly sure” is another somewhat outdated way of identifying the phenomenon of legislative repetitiveness, with courts appreciating rather than merely excoriating the reality.
  6. [6]. See, e.g., Gustafsonv.Alloyd Co., Inc.,513 U.S. 561, 574 (1995) (noting that courts should tend to avoid interpretations of statutes that “render[] some words altogether redundant”). In a recent study, John Golden found that the rule against superfluities is massively popular with courts—and that other textual canons “lag far behind.” John M. Golden, Redundancy: When Law Repeats Itself, 94 Tex. L. Rev. 629, 653–54 (2016); see also Anita S. Krishnakumar, Statutory Interpretation in the Roberts Court’s First Era: An Empirical and Doctrinal Analysis, 62 Hastings L.J. 221, 243 n.100 (2010) (acknowledging the continuing force of the rule against redundancies on the modern Supreme Court).

              For criticisms of the rule against superfluities, see, e.g., Richard A. Posner, The Federal Courts: Crisis and Reform 281 (1985) (“The conditions under which legislators work are not conducive to careful, farsighted, and parsimonious drafting. Nor does great care guarantee economy of language; a statute that is the product of compromise may contain redundant language as a by-product of the strains of the negotiating process.”); Richard A. Posner,
  7. [7]. See Karl N. Llewellyn, Remarks on the Theory of Appellate Decision and the Rules or Canons About How Statutes Are to Be Construed, 3 Vand. L. Rev. 395, 401–06 (1950). It is not clear why the other Latin canons like ejusdem generis, noscitur a sociis, and expressio unius have remained at the tip of judicial tongues but ex abundanti cautela has mostly faded. The sheer power of the anti-redundancy canon is evidenced in a recent study of judicial uses from 2000 to 2015. It was deployed around 4291 times, as compared with 458 deployments of ejusdem generis, 296 deployments of noscitur a sociis, and 991 uses of expressio unius. See Golden, supra note 6, at 653 n.94, 654 n.97. Apparently, ex abundanti cautela was not part of Golden’s calculus. But the overwhelming robustness of anti-redundancy in statutory interpretation may explain, in part, why its natural counter-canon has faded from view.

  8. [8]. See generally Anita S. Krishnakumar, Dueling Canons, 65DukeL.J. 909 (2016).

  9. [9]. What we refer to as the rule against superfluities or the anti-surplusage canon may be manifested in corollary framings as a rule to avoid redundancy or as a part of the “whole act rule.” The logic of the rule also undergirds the presumption of meaningful variation: A change of wording or phrasing denotes a change of meaning. See William N. Eskridge, Jr., Philip P. Frickey, Elizabeth Garrett & James J. Brudney, Cases and Materials on Legislation and Regulation: Statutes and the Creation of Public Policy 675–79 (5th ed. 2014).

  10. [10]. See, e.g., NLRB v. Sw. Gen., Inc., 137 S. Ct. 929, 941 (2017) (Roberts, C.J.) (describing superfluities as “a result we typically try to avoid”); Roberts v. Sea-Land Servs., Inc., 566 U.S. 93, 103 (2012) (Sotomayor, J.) (refusing to construe text “in a manner that renders it ‘entirely superfluous in all but the most unusual circumstances’” (quoting TRW Inc. v. Andrews, 534 U.S. 19, 29 (2001))); Corley v. United States, 556 U.S. 303, 314 (2009) (Souter, J.) (referring to rule against superfluities as “one of the most basic interpretive canons”).

  11. [11]. Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 174–79 (2012). William Eskridge’s counter-text also discusses what he calls the “[a]nti-[s]urplusage (-[r]edundancy) [c]anon[,]” though he concedes that “the premise upon which this canon rests is often unrealistic from the point of view of the operation of the legislative process outside of the drafting office.” William N. Eskridge, Jr., Interpreting Law: A Primer on How To Read Statutes and the Constitution 112–14 (2016).

  12. [12]. See Abbe R. Gluck & Lisa Schultz Bressman, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901, 934 (2013).

  13. [13]. See id. at 932. Noscitur a sociis translates as “it is known from its associates,” indicating that the meaning of an ambiguous word may be clarified by reference to words associated with it, usually in a list. Expressio unius est excusio alterius translates as “the expression of one thing excludes the other;” it is invoked to conclude that items not included in a specific provision are assumed not to be covered by the statutory phrase.

  14. [14]. See Golden, supra note 6, at 653 n.94, 654 n.97.

  15. [15]. See Jacob Scott, Codified Canons and the Common Law of Interpretation, 98 Geo. L.J. 341, 368 (2010). Scott identifies Colorado, Iowa, Minnesota, Montana, New Mexico, North Dakota, Ohio, Oregon, Pennsylvania, and Texas. Id. at 415 (citing state statutes). We confirmed his data on those states in 2019.

  16. [16]. Scalia & Garner, supra note 11, at 177.

  17. [17]. Fort Stewart Sch. v. FLRA, 495 U.S. 641, 646 (1990).

  18. [18]. See H.R. Rep. No. 107-170, at 48–49 (2001) (referencing Representative Frank’s comments).

  19. [19]. See Gluck & Bressman, supra note 12, at 934–35 (discussing why drafters “are often purposefully redundant to satisfy audiences other than courts”).

  20. [20]. Of course, as with any canon, the disfavoring of surplusage is not an absolute rule. See Lamie v. U.S. Trustee, 540 U.S. 526, 536 (2004) (“[O]ur preference for avoiding surplusage constructions is not absolute.”); Chickasaw Nation v. United States, 534 U.S. 84, 94 (2001) (suggesting that the disfavor “is sometimes offset by the canon that permits a court to reject words ‘as surplusage’ if ‘inadvertently inserted or if repugnant to the rest of the statute’”). However, as far as canons go, this rule is very robust. See Golden, supra note 6, at 653–54. So robust that the rule’s counter-canon has faded from view—until now.

  21. [21]. National Labor Relations Act, 29 U.S.C. § 160(c) (2012).

  22. [22]. See 18 U.S.C. § 2255(a), discussed in Doe v. Boland, 698 F.3d 877, 880–82 (6th Cir. 2012).

  23. [23]. 11 U.S.C. § 363(f).

  24. [24]. See, e.g., State v. Nelson, 842 N.W.2d 433, 443–44 (Minn. 2014) (construing failure to provide “care and support” in state child support statute), discussed infra Section II.C.

  25. [25]. Thanks to Aaron Bruhl for this nuance.

  26. [26]. See Yates v. United States, 135 S. Ct. 1074, 1084 (2015) (quoting 18 U.S.C. § 1512(c)(1) (2015)). This was Justice Kagan’s position about what Congress did in Yates v. United States. See id. at 1093–98 (Kagan, J., dissenting); supra note 1 and accompanying text.

  27. [27]. See 11 U.S.C. §§ 546(e), and 741(7) (2012), discussed in In re Lehman Bros. Holdings Inc., 469 B.R. 415, 437–38 (Bankr. S.D.N.Y. 2012). For an analogous state law illustration, Hawaii’s smoke-free public places workplaces law doubles down in its definitions and prohibition sections in identifying areas “open to the public” where smoking is prohibited. See Haw. Rev. Stat.
    §§ 328J-1, 328J-3 (2010).

  28. [28]. See, e.g., Federal Pesticide Act of 1978, 7 U.S.C. § 136h(g)(3) (2012) (restricting the EPA Administrator’s freedom to disclose certain information to foreign or multinational pesticide producers while also expressly providing that the criminal penalty provisions of 18 U.S.C. § 1001 apply).

  29. [29]. See generally United States v. Hansen, 772 F.2d 940, 946–47 (D.C. Cir. 1985) (Scalia, J.) (reasoning that Congress sometimes includes in new laws a reference to earlier enacted text such as “[18 U.S.C.] § 1001 as a means of reminding those subject to the new laws of the self-operative, previously enacted sanctions, or as a means of clarifying for its own Members who voted upon the new laws the consequences of their action.” (emphasis added)).

  30. [30]. See, e.g., amendment to 35 U.S.C. §§ 303(a), and 312(a) (2012), discussed infra Section II.B.3.

  31. [31]. Francis Lieber, Legal and Political Hermeneutics, or Principles of Interpretation and Construction in Law and Politics, with Remarks on Precedents and Authorities 30 (1839). Thanks to Peter Strauss for reminding us of this discussion.

  32. [32]. We discuss examples of this approach infra Sections II.B and II.C.

  33. [33]. See Victims of Trafficking and Violence Protection Act of 2000, Pub. L. No. 106-386,
    § 2004, 114 Stat. 1464, 1546 (codified as amended at 22 U.S.C. § 7101 (2012)) (amending 22 U.S.C. § 122b with the Twenty-First Amendment Enforcement Act). Initially this provision was part of H.R. 2031, the Twenty-First Amendment Enforcement Act.

  34. [34]. Internet Tax Freedom Act, 47 U.S.C. § 151 note (2012) (prohibiting taxation and certain regulation of internet commerce). For floor discussion, see 145 Cong. Rec. H6856, H6868, H6870 (daily ed. Aug. 3, 1999).

  35. [35]. 145 Cong. Rec. H6868 (introducing amendment to provide inter alia that “[n]othing in this Act may be construed to modify or supersede the operation of the Internet Tax Freedom Act”).

  36. [36]. 145 Cong. Rec. H6870.

  37. [37]. Id.

  38. [38]. See H.R. 2883, 115th Cong. (1st Sess. 2017), discussed at 163 Cong. Rec. H6010­–23 (daily ed. July 19, 2017).

  39. [39]. 163 Cong. Rec. H6016, H6019 (daily ed. Aug. 3, 1999).

  40. [40]. 163 Cong. Rec. H6016.

  41. [41]. 163 Cong. Rec. H6020 (daily ed. Aug. 3, 1999).

  42. [42]. 163 Cong. Rec. H6020, H6023 (daily ed. Aug. 3, 1999).

  43. [43]. See H.R. Rep. No. 107-120, at 2 (1st Sess. 2001) (accompanying H.R. 1866, 107th Cong., 116 Stat. 1758 (1st Sess. 2001) (enacted)).

  44. [44]. 35 U.S.C. § 303(a) (2012); see also id. § 312(a) (listing the requirements for a petition to institute an inter partes review of a patent).

  45. [45]. See In re Portola Packaging, Inc., 110 F.3d 786, 787 (Fed. Cir. 1997). For a discussion of the case see H.R. Rep. No. 107-120, at 2.

  46. [46]. See H.R. Rep. No. 107-120, at 2.

  47. [47]. 21st Century Department of Justice Appropriations Authorization Act, Pub. L. No. 107-273, § 13105, 116 Stat. 1758, 1900 (2002) (codified as amended at 35 U.S.C. §§ 303(a), 312(a) (2012)).

  48. [48]. H.R. Rep. No. 107-120, at 13 (statement of Rep. Coble).

  49. [49]. Id. at 2.

  50. [50]. H.R. 760, 103d Cong. (1993).

  51. [51]. See Biotechnology Patent Protection Act of 1991, H.R. 1417, 102d Cong.; Biotechnology Patent Protection Act of 1991: Hearing on H.R. 1417 Before the House Judiciary Subcomm. on Intell. Prop. & Judicial Admin. of the H. Comm. on the Judiciary, 102d Cong. 65 (1991).

  52. [52]. See Amending Title 35, United States Code, With Respect to Patents on Certain Processes: Hearing on H.R. 760 Before the Subcomm. on Intell. Prop. & Judicial Admin. of the H. Comm. on the Judiciary, 103d Cong. 1 (1993) (statement of Chairman Hughes).

  53. [53]. See id. at 38,­ 44 (prepared statement of George W. Ebright on behalf of the Indus. Biotechnology Ass’n).

  54. [54]. See id. at 4, 24–27, 44 (amending 35 U.S.C. § 103 to overturn In re Durden, 763 F.2d 1406 (Fed. Cir. 1985)).

  55. [55]. See id. at 6, 44 (amending 35 U.S.C. § 271).

  56. [56]. Id. at 38, 44 (statement of George W. Ebright).

  57. [57]. Id. at 30 (statement of Michael K. Kirk).

  58. [58]. Id. (exchange between Rep. Reed and Michael K. Kirk).

  59. [59]. Id. at 52 (exchange between Rep. Hughes and Steven M. Odre).

  60. [60]. See Golden, supra note 6, at 670, 673–99.

  61. [61]. Id. at 699.

  62. [62]. The bill was reported by the Judiciary Committee on July 27, 2001. See H.R. Rep. No. 107-170, at 1 (2001). The bill passed the House four days later on July 31. Numerous other unsuccessful efforts to enact such a ban occurred between 1998 and 2009. See Part Four: Cloning Policy in the United States, New Atlantis, Summer 2015, at 74–79.

  63. [63]. See H.R. Rep. No. 107­-170, at 46 (2001) (citing the Lofgren-Conyers amendment).

  64. [64]. See id. at 47–48 (exchange between Rep. Smith and Rep. Lofgren).

  65. [65]. Id. at 48.

  66. [66]. See id. at 48–49. The amendment was ultimately defeated in committee by a vote of
    18–11. See id. at 51. We are not sure why Frank thinks lawyers created the idea of “belt-and-suspenders.” We traced British uses of “belt-and-braces” and American uses of “belt-and-suspenders” that pre-date any legal references. See, e.g., Belt, The Oxford English Dictionary, Phrase 4 (“belt and braces (originally and chiefly British). Later also belt and suspenders” (citing Robert Macintosh & Mary Ostlere, Local Analgesia Head & Neck xvii, 131 (1955) and T.W. McRae, Impact Computers on Accounting vi. 187 (1965))); Why be an Optimist?, 92 J. Educ. 634, 634 (1920) (“The president of a university is said to have likened a pessimist to a man who wears both a belt and suspenders and a county superintendent replied that an optimist is one who tries to get along without either.”); Calvin I. Ryan, It’s Anxiety, Not Optimism That Wins!, 34 Rotarian 29, 29 (1929) (“A pessimist may wear both belt and suspenders.”); Sunday Pleads for Old Creed, L.A. Times, Nov. 21, 1932, at A2 (“A pessimist is a man who wears both a belt and suspenders.”).

  67. [67]. See S. 2663, 110th Cong. 2d Session, § 31 (2008) (enacted).

  68. [68]. See 154 Cong. Rec. S1581–82 (2008) (amendment from Sen. DeMint).

  69. [69]. Id. at S1582.

  70. [70]. Id.

  71. [71]. See id.; see also id. at S15834 (remarks of Sen. Durbin, opposing the amendment and citing the belt-and-suspenders justification). The amendment was later withdrawn without a vote. See id. at S1584.

  72. [72]. H.R. Rep. No. 110-787, at 75–77 (2008) (Conf. Rep.). The Act itself is the Consumer Product Safety Improvement Act of 2008, Pub. L. No. 110-314, 122 Stat. 3016 (codified as amended at 15 U.S.C. § 2051 note (2012)).

  73. [73]. For some other examples, see 158 Cong. Rec. S1978 (2012) (remarks of Sen. Leahy); 150 Cong. Rec. S11852 (2004) (remarks of Sen. Cornyn); 141 Cong. Rec. S9673, S9677–78 (1995) (remarks of Sens. Kyl and Levin); J.W. Verret, Treasury Inc.: How the Bailout Reshapes Corporate Theory and Practice, 27 Yale J. on Reg. 283, 307–12 (2010) (discussing how the statute creating the Troubled Asset Relief Program (“TARP”) bailout added several “suspenders” of protection from liability for the U.S. Treasury in addition to the already existing “belt” of the principle of sovereign immunity itself).

  74. [74]. See Scott, supra note 15, at 415 (listing state statutes that have enacted the rule against surplusage). Both Connecticut and Texas have codified a wide variety of canons. See id. at 411
    –26. As it happens, however, courts do not always feel they need to abide by legislatively-enacted interpretive directions, so identifying codified canons does not always give insight into the law that is actually applied in the courts. See, e.g., State v. Pena, 683 P.2d 744, 748–49 (Ariz. Ct. App. 1983), aff’d,683 P.2d 743 (Ariz. 1984)(applying rule of lenity notwithstanding the legislative abrogation by statute). Scott opines that judges might ignore legislative direction about interpretation because they believe that “legislative control over judicial interpretive methodology is unconstitutional.” Scott, supra note 15, at 410. Yet, as we will show infra, courts do sometimes follow legislative direction, too.

  75. [75]. S. Res. 910, 2013 Gen. Assemb., Reg. Sess. (Conn. 2013) (outlining Conn. Pub. Act No. 13-176).

  76. [76]. See id. § 1.

  77. [77]. See Transcript of Connecticut House of Representatives at 65 (May 23, 2013) (summarizing the amendment offered by Rep. Smith).

  78. [78]. Id.

  79. [79]. See id. at 66. The bill as amended passed the House on May 23 and became law on October 1, 2013. See Conn. S. Res. 910.

  80. [80]. See Transcript of Connecticut Senate at 16 (May 25, 2017) (remarks of Sen. Boucher, Transportation Committee Chair).

  81. [81]. See id. at 18–19 (remarks of Sen. Boucher).

  82. [82]. Id. at 21–22 (remarks of Sen. Suzio).

  83. [83]. See S. 76, 2017 Gen. Assemb., Reg. Sess. (Conn. 2017) (outlining Public Act No. 17-174, approved July 11, 2017).

  84. [84]. Transcript of Connecticut Senate, at 34 (May 25, 2017) (remarks of Sen. Looney).

  85. [85]. S. 302, 85th Leg., Reg. Sess. (Tex. 2017) (adding section 81.062 “State Bar Admission and Religious Belief” to the Texas Government Code).

  86. [86]. See Tex. Civ. Prac. & Rem. Code Ann. § 110.003 (2019).

  87. [87]. H. Journal Supplement, 85th Leg., Reg. Sess., at S302 (Tex. May 15, 2017).

  88. [88]. Id.

  89. [89]. Id. at S303.

  90. [90]. When Turner asked incredulously if Leach believed there was a risk that the Texas Supreme Court would in the future, look to the ABA over the U.S. Constitution, Leach replied, “I do believe there’s that risk. Stranger things have happened.” Id.

  91. [91]. See H.B. 82-15, Reg. Sess., at 4 (Tex. 2011) (amending section 171.002, Health and Safety Code).

  92. [92]. See H. Journal Supplement, 82d Leg., Reg. Sess., at S76–S77 (Tex. March 3, 2011).

  93. [93]. Id. at S77.

  94. [94]. See Cass R. Sunstein, After the Rights Revolution: Reconceiving the Regulatory State 147 (1993) (referring to canons as “background principles of interpretation that are used in statutory construction”); William N. Eskridge, Jr. & Philip P. Frickey, Foreword: Law as Equilibrium, 108 Harv. L. Rev. 26, 65–71 (1994) (same). For a more restrictive definition of what should qualify as a canon, see Anita S. Krishnakumar & Victoria F. Nourse, The Canon Wars, 97 Tex. L. Rev. 163, 181–90 (2018) (identifying as potential tests or measures the frequency of Supreme Court use; longevity of the rule; and justifications for the rule, while eschewing a requirement that the Court declare a canon’s specific existence).

  95. [95]. See supra Section II.A.

  96. [96]. See supra notes 4–5 and accompanying text.

  97. [97]. We refer here to the Krishnakumar and Nourse criteria, supra note 94, at 181–90.

  98. [98]. For an early effort to draw attention to the ways the judicial “interpretive” virtue of clarity in drafting stood in tension with other important “constitutive” legislative values of action and agreement, see generally Victoria A. Nourse & Jane S. Schacter,The Politics of Legislative Drafting: A Congressional Case Study, 77 N.Y.U. L. Rev. 575 (2002) (conducting a case study of drafting by the Senate Judiciary Committee).

  99. [99]. Scalia & Garner, supra note 11, at 59.

  100. [100]. See infra Section III.A.

  101. [101]. See infra Section III.B.

  102. [102]. See infra Section III.C.

  103. [103]. See infra Part IV.

  104. [104]. Freeman v. Quicken Loans, Inc., 566 U.S. 624, 635 (2012).

  105. [105]. Id.; see also 12 U.S.C. § 2607(b) (2012) (“Splitting charges”).

  106. [106]. 12 U.S.C. § 2607(b).

  107. [107]. Freeman, 566 U.S. at 635.

  108. [108]. Id.

  109. [109]. See also Ransom v. FIA Card Servs., N.A., 562 U.S. 61, 81 (2011) (Scalia, J., dissenting) (“The canon against superfluity is not a canon against verbosity. When a thought could have been expressed more concisely, one does not always have to cast about for some additional meaning to the word or phrase that could have been dispensed with. This has always been understood. A House of Lords opinion holds, for example, that in the phrase ‘“in addition to and not in derogation of”’ the last part adds nothing but emphasis.” (quoting Davies v. Powell Duffryn Associated Collieries, Ltd., [1942] A. C. 601, 607)). Why is Scalia citing the House of Lords here, when surely there are some domestic statutes to make the point? Thanks to Anita Krishnakumar for the observation.

  110. [110]. United States v. Bronstein, 849 F.3d 1101 (D.C. Cir. 2017).

  111. [111]. 40 U.S.C. § 6134.

  112. [112]. See United States v. Bronstein, 151 F. Supp. 3d 31, 41–44 (D.D.C. 2015).

  113. [113]. Bronstein, 849 F.3d at 1108.

  114. [114]. Id. at 1108–09.

  115. [115]. Id. at 1110.

  116. [116]. Id. (alteration in original) (quoting Scalia & Garner, supra note 11, at 176–77).

  117. [117]. The court here also quotes Scalia & Garner calling the practice of “stringing out synonyms and near-synonyms” “retrograde.” Id. (quoting Scalia & Garner, supra note 11, at 179). Notwithstanding all this name-calling, Scalia and Garner should be commended for admitting that the anti-surplusage canon does commonly need to give way to what we are calling the belt-and-suspenders canon, which acknowledges real drafting practices.

  118. [118]. Id. at 1110.

  119. [119]. Id.

  120. [120]. See Navajo Nation v. Azar, 302 F. Supp. 3d 429, 439 n.7 (D.D.C. 2018).

  121. [121]. See also Doe v. Boland, 698 F.3d 877, 881 (6th Cir. 2012) (holding that the civil remedial scheme in 18 U.S.C. § 2255 for damages for “victims” of child pornography does not additionally require plaintiffs to show “personal injury” because “the presumption against surplusage does not apply to doublets—two ways of saying the same thing that reinforce its meaning.”); id. (“The U.S. Code is replete with meaning-reinforcing redundancies: an invalid contract is ‘null and void’; agency action must not be ‘arbitrary and capricious’; bureaucrats send ‘cease and desist’ letters; a bankruptcy trustee can sell a debtor’s property ‘free and clear’ of other interests; and so on. When faced with an agency order, to use one of these examples, how could a citizen cease but not desist? He could not.” (citations omitted)).

  122. [122]. See, e.g., McEvoy v. IEI Barge Servs. Inc., 622 F.3d 671, 677 (7th Cir. 2010) (“The fact that the different subparagraphs of § 7604(f)may overlap to a degree is no reason to reject the natural reading of a statute. Congress may choose a belt-and-suspenders approach to promote its policy objectives, and it appears that this is what it was doing when it added the broader provision to the statute through a later amendment.”).

  123. [123]. Atl. Fish Spotters Ass’n v. Evans, 321 F.3d 220, 226 (1st Cir. 2003) (“It is certainly plausible ... that Congress intended the phrase ‘[n]one of the funds provided in this or any previous Act, or hereinafter made available to the Department of Commerce’ as nothing more than a particularly emphatic way of stating the stock phrase ‘this or any other act,’ which appears regularly in appropriations bills... . We think it not unlikely that Congress used the belt-and-suspenders version of this trite phrase ... . Since the rudimentary phrase fails to overcome the presumption against permanence in appropriations bills, the rephrasing of it, with only inconsequential variations, should not be deemed to establish permanence.” (citation omitted)).

  124. [124]. Rajala v. Gardner, 709 F.3d 1031, 1038 (10th Cir. 2013).

  125. [125]. 11 U.S.C. § 362(a)(3) (2012).

  126. [126]. Id. § 541(a)(1), (a)(3).

  127. [127]. Id. §§ 550(a), 548.

  128. [128]. Rajala, 709 F.3d at 1037–38.

  129. [129]. Id. at 1038 (alteration in original) (quoting TRW Inc. v. Andrews, 534 U.S. 19, 31 (2001)).

  130. [130]. Id. (citing FDIC v. Hirsch (In re Colonial Realty Co.), 980 F.2d 125, 131 (2d Cir. 1992)).

  131. [131]. Id. at 1038–39.

  132. [132]. Id. at 1038.

  133. [133]. See id. at 1039 (“Both sides present plausible arguments regarding Congress’s intent [contrasting H.R. Rep. No. 595 (1978) with the structure of the statute]. Therefore, the plain meaning of the statutory language should control.”).

  134. [134]. See id. at 1038­–39.

  135. [135]. Compare Am. Nat’l Bank of Austin v. MortgageAmerica Corp. (In re MortgageAmerica Corp.),714 F.2d 1266, 1275(5th Cir. 1983) (holding allegedly fraudulently conveyed assets to be subject to the automatic stay because one could say that the debtor retains “equitable interest” in the assets under a literal reading of § 541(a)(1)), with In re Colonial Realty Co., 980 F.2d at 131 (holding like Rajala that because section § 541(a)(3) expressly provides that estate property “includes ‘[a]ny interest in property that the trustee recovers’ under ... § 550[,]” the automatic stay does not apply until the transfer is avoided and recovered).

  136. [136]. Babbitt v. Sweet Home Chapter of Cmtys. for a Great Or., 515 U.S. 687 (1995).

  137. [137]. Id. at 691; see also 50 C.F.R. § 17.3 (2018) (further defining through a regulation the statutory terms “take” and “harm”).

  138. [138]. Endangered Species Act of 1973, 16 U.S.C. § 1532(19) (2012). It is plausible to argue that harass and kill—like harm—can cover indirect or unintended injuries. Still, an initial focus of the majority and dissent was on the meaning of “harm” in the midst of this string of apparent synonyms and near-synonyms.

  139. [139]. Babbitt, 515 U.S. at 720 (Scalia, J., dissenting).

  140. [140]. Sweet Home Chapter of Cmtys. for a Great Or. v. Babbitt, 17 F.3d 1463, 1464 (D.C. Cir. 1994).

  141. [141]. Babbitt, 515 U.S. at 698 (“A reluctance to treat statutory terms as surplusage supports the reasonableness of the Secretary’s interpretation.”).

  142. [142]. Id. at 697–701.

  143. [143]. Id. at 698–701.

  144. [144]. State v. Nelson, 842 N.W.2d 433 (Minn. 2014).

  145. [145]. See id. at 435 (discussing Minn. Stat. § 609.375 subds. 1, 2a(1) (2012)).

  146. [146]. Nelson, 842 N.W.2d at 435.

  147. [147]. Id.

  148. [148]. Id. at 440­–43.

  149. [149]. Id. at 442–44.

  150. [150]. Id. at 442. Relatedly, the majority emphasized that other subdivisions of the “care-and-support statute” used variations on the term “support” (without “care”) to refer to a person’s financial obligations to a spouse or child. See id. at 439.

  151. [151]. Id. at 444.

  152. [152]. See Scott, supra note 15, at 415 (citing Minn. Stat. §§ 645.16–.17 (2008)).

  153. [153]. Nelson, 842 N.W.2d at 447 (Dietzen, J., dissenting) (“providing that ‘[e]very law shall be construed, if possible, to give effect to all its provisions’” (citingMinn. Stat. § 645.16 (2012))). As we highlighted above, however, interpretive directions by legislatures do not always get warm welcomes from the judiciary. See supra note 74 and accompanying text. Another example of this phenomenon had the Wisconsin Supreme Court invoking titles of statutes, notwithstanding that the Wisconsin legislature had passed a legislative direction that “titles to subchapters, sections, subsections, paragraphs and subdivisions of the statutes ... are not part of the statutes.” Wisc. Stat. Ann. 990.001(6) (West 2007); see State v. Matasek, 846 N.W.2d 811, 818 & n.22 (Wisc. 2014) (utilizing a title to help resolve a statutory interpretation issue); Rouse v. Theda Clark Med. Ctr., Inc., 735 N.W.2d 30, 37 (Wisc. 2007) (same).

  154. [154]. The dissent also invoked precedent, citing “over 90 years” of construing the phrase “care and support” to mean no more than “the ... obligation to provide financial support;” as well as context, arguing that “‘care and support’ ... ‘to a spouse or child’ ... renders unreasonable” a construction that imposes a legal duty “to provide psychological support and a nurturing environment to one’s estranged spouse.” Nelson, 842 N.W.2d at 445–78 (Dietzen, J., dissenting).

  155. [155]. Freeman v. Quicken Loans, Inc., 566 U.S. 624, 635 (2012).

  156. [156]. Nelson, 842 N.W.2d at 448 (Dietzen, J., dissenting).

  157. [157]. Id. at 449 (citing Bryan A. Garner, The Redbook: A Manual on Legal Style § 11.2(f) (2d ed. 2006)).

  158. [158]. Id.

  159. [159]. See Elizabeth Ahlin, Lawmakers Step in on Minnesota Child Support Issue, Minn. Law., Feb. 27, 2014.

  160. [160]. Scalia & Garner, supra note 11, at 179.

  161. [161]. United States v. Smith, 756 F.3d 1179, 1195–96 (10th Cir. 2014).

  162. [162]. Id. at 1180.

  163. [163]. 18 U.S.C. § 1951 (2012); Smith, 756 F.3d at 1180–81.

  164. [164]. Smith, 756 F.3d at 1180–81 (alteration in original) (quoting 18 U.S.C. § 924(c)).

  165. [165]. Id.

  166. [166]. Id. at 1184.

  167. [167]. See id. at 1185­–87 (quoting 18 U.S.C. § 1028A).

  168. [168]. Id. at 1185–86 (citing 18 U.S.C. § 1028A(a), (b)(2)).

  169. [169]. Id. at 1186 (quoting 18 U.S.C. § 1028A(b)(3)).

  170. [170]. Id. at 1186–87.

  171. [171]. Id. at 1196 (Lucero, J., dissenting) (alteration in original) (citations omitted) (quoting H.R. Rep. No. 108-528, at 10 (2004), reprinted in 2004 U.S.C.C.A.N. 779, 785–86).

  172. [172]. Intriguingly, then-Judge Scalia prevailed on a similar argument two decades earlier. See United States v. Hansen, 772 F.2d 940, 947 (D.C. Cir. 1985) (“Even if one thinks the [redundant] references were included for their operative effect, they necessarily establish no more than that Congress chose in some cases to make assurance doubly sure, but did not do so here.”).

  173. [173]. Smith, 756 F.3d. at 1187 (quoting Duncan v. Walker, 533 U.S. 167, 174 (2001)).

  174. [174]. 42 U.S.C. §§ 1973gg-2(a)(2), -4(a)(1)–(2) (2012).

  175. [175]. Gonzalez v. Arizona, 677 F.3d 383, 398 (9th Cir. 2012).

  176. [176]. See id. at 398–401.

  177. [177]. See id. at 439–42 (Kozinski, C.J., concurring).

  178. [178]. See id. at 439 (Kozinski, C.J., concurring) (suggesting that the broad or inclusive construction “wouldn’t render the federal form superfluous” but instead would amount to “wearing a belt and suspenders”).

  179. [179]. 139 Cong. Rec. 5099 (March 16, 1993) (remarks of Sen. Ford).

  180. [180]. See id.

  181. [181]. H.R. Rep. No. 103-66, at 23–24 (1993) (Conf. Rep.), reprinted in 1993 U.S.C.C.A.N. 140, 148–49.

  182. [182]. Id. This language is quoted in Gonzalez, 677 F.3d at 441 (Kozinski, C.J., concurring).

  183. [183]. Gonzalez, 677 F.3d at 441(Kozinski, C.J., concurring). Kozinski went on to explain that in both chambers, a small number of legislators objected to the elimination of the Simpson amendment as encouraging voter fraud, but their arguments were rebuffed by the majority. Id.

              The prospects for a belt-and-suspenders approach to another provision of the NVRA
  184. [184]. See supra Section II.A.2.

  185. [185]. See supra Section II.A.1.

  186. [186]. For an argument that courts should be more regularly “subservien[t] ... to legislative instruction,” see Peter L. Strauss, Statutes That Are Not Static—The Case of the APA, 14 J. Contemp. Legal Issues 767, 777 (2005).

  187. [187]. Scalia & Garner, supra note 11, at 179.

  188. [188]. Id.

  189. [189]. Id. at 177.

  190. [190]. Id.

  191. [191]. Id.

  192. [192]. Id. at 179.

  193. [193]. Id.

  194. [194]. See Brett M. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2122 (2016) (reviewing Robert A. Katzmann, Judging Statutes (2014)); Seven-Sky v. Holder, 661 F.3d 1, 38 (D.C. Cir. 2011) (Kavanaugh, J., dissenting); Gonzalez v. Arizona, 677 F.3d 383, 441 (9th Cir. 2012) (Kozinski, C.J., concurring).

  195. [195]. For a discussion of this feature of Nelson, see supra note 153 and accompanying text. For a discussion of the ways judges sometimes ignore legislative interpretive directions, see supra note 74 and accompanying text.

  196. [196]. In his Gonzalez concurrence, Judge Kozinski makes this very point, observing that even staunchly textualist justices might well not object to the use of legislative history “with equal vigor where, as here, the statutory language is in equipoise [between belt-and-suspenders and surplusage] and both chambers affirmatively rejected efforts to authorize precisely what Arizona is seeking to do.” Gonzalez, 677 F.3d at 442 (Kozinski, C.J., concurring).

  197. [197]. See supra text accompanying notes 138–52.

  198. [198]. Scalia & Garner, supra note 11, at 179.

  199. [199]. For discussion about the “ordering problem” with ranking canons, see Krishnakumar & Nourse, supra note 94, at 168–74.

  200. [200]. See supra text accompanying notes 153–63.

  201. [201]. See, e.g., Moses Maimonides, The Guide for the Perplexed pt. 3, ch. 50 (1190); Sanhedrin 99b. Even as a principle of biblical interpretation, there is plenty of debate about how to think about style, doublets, and belt-and-suspenders drafting techniques in the Hebrew Bible. See, e.g., Are There Unnecessary Words in Torah?, Mi Yodeya,
    questions/18799/are-there-unnecessary-words-in-torah [].


Ethan J. Leib is the John D. Calamari Distinguished Professor of Law at Fordham Law School.


James Brudney holds the Joseph Crowley Chair in Labor and Employment Law at Fordham Law School.

We thank Aaron Bruhl, Rick Hills, Olati Johnson, Anita Krishnakumar, Mark Levin, Vic Nourse, Richard Re, and Peter Strauss for their thoughtful comments on prior drafts. We are grateful to Thomas Lloyd, Janet Kearney, Gail McDonald, and Alison Shea for excellent research assistance, and to Fordham Law School for generous financial support.