This dispute centers on what constitutes an employee pension benefit plan under the Employee Retirement Income Security Act (ERISA), and its resolution determines whether the duties and protections of ERISA apply to the plan at issue. Plaintiff Dan Wilson, the former President and Chief Executive Officer of Defendant Safelite Group, Inc., sued Safelite for breach of contract and negligent misrepresentation arising from the company's alleged mismanagement of its deferred compensation plan for executive employees. Finding that the plan was an employee pension benefit plan under 29 U.S.C. § 1002 (2)(A)(ii) and not a bonus plan exempted from ERISA under 29 C.F.R. § 2510.3-2 (c), the district court granted Safelite's motion for partial summary judgment on the basis that Wilson's state law claims were preempted by ERISA. Based on the plain text of ERISA and the bonus plan regulation, we AFFIRM .

I. OVERVIEW

Wilson was the President and CEO of Safelite from 2003 to 2008. In 2005, Safelite's Board of Directors created the Safelite Group, Inc. 2005 Transaction Incentive Plan (TIP), which provided for substantial bonus payments to its participants-five *432 Safelite executives, including Wilson-if they secured a strategic buyer for the company.

By late 2006, Belron SA emerged as a likely buyer. Realizing that Belron's acquisition of Safelite would trigger significant payments under the TIP that could increase participants' tax obligations, on December 29, 2006, the Board adopted the Safelite Group, Inc. Nonqualified Deferred Compensation Plan (Safelite Plan), a plan to allow participants to defer compensation and thereby avoid certain tax consequences. At the time the Safelite Plan was adopted, only four executive employees, including Wilson, were eligible to participate in it. In February 2007, less than two months later, Belron purchased Safelite for $334 million, generating substantial payments to the TIP participants that could be deferred by operation of the Safelite Plan.

The Safelite Plan, the plan at issue, allows eligible executive employees to defer two types of income: (1) compensation (defined as a participant's base annual salary and any annual or long-term bonuses) and (2) "TIP Amounts" triggered by Safelite's sale to Belron. To become a participant in the Safelite Plan, an eligible employee must submit a completed election form and must indicate what income to defer. To defer any TIP Amount, an eligible employee was required to submit an election form on or before December 31, 2006. An eligible employee could choose to defer each upcoming year's compensation by submitting an election form before January 1 of that year. The election forms provide spaces to indicate what percentage of the employee's TIP Amount, base annual salary, and bonus he seeks to defer and in which year or years the employee wishes to receive distributions of his deferred income, either during employment or after the termination of employment.

The Safelite Plan provides two timing options for participants to receive distributions of deferred income. The default distribution of deferred compensation for each participant is payment "in a lump sum as soon as administratively feasible after the [participant] Terminates" from employment. A participant can elect to receive deferred distributions on January 1 of a designated year or following a disability, and that distribution can be "in a lump sum or monthly ... or annual payments over a period of up to ten years," subject to some limitations. Distributions made during a participant's employment are referred to as "in-service distributions."

Wilson properly submitted an election form and so became a participant in the Safelite Plan. Between 2006 and 2013, he elected to defer hundreds of thousands of dollars of compensation each year. Wilson left Safelite on July 5, 2008. By 2014, Wilson had deferred compensation totaling $9,111,384. That year, a federal audit revealed that some of Wilson's elections failed to comply with 26 U.S.C. § 409A, a tax statute regulating deferred compensation plans. As a result, Wilson owed income taxes and incurred substantial tax penalties.

On September 12, 2016, Wilson sued Safelite in federal court, asserting state law claims for breach of contract and negligent misrepresentation. Safelite moved for partial summary judgment on Wilson's state law claims, arguing that they were preempted by ERISA. The district court granted Safelite's motion, finding that the Safelite Plan met the statutory definition of "employee pension benefit plan" under ERISA Section 3(2)(A)(ii), 29 U.S.C. § 1002 (2), and was not a bonus plan exempted from ERISA coverage under Department of Labor (DOL) regulation *433 29 C.F.R. § 2510.3-2 (c). 1 The district court granted Wilson 28 days to file an amended complaint asserting claims under ERISA's civil enforcement provision. Wilson chose not to amend his complaint, and the district court entered final judgment on April 19, 2018. Wilson timely appealed.

II. ANALYSIS

We review a grant of summary judgment de novo. Kolkowski v. Goodrich Corp. , 448 F.3d 843 , 847 (6th Cir. 2006). Whether an ERISA plan exists is usually "a question of fact to be answered in light of all the surrounding circumstances and facts" and reviewed for clear error. Id. at 847-48 . But "[w]here key facts are undisputed, the determination is better considered a mixed question of law and fact that we review de novo." Wolf v. Causley Trucking, Inc. , 719 F. App'x 466 , 470 n.1 (6th Cir. 2017). The standard of review in a mixed question of law and fact generally depends on "whether answering it entails primarily legal or factual work." U.S. Bank Nat. Ass'n v. Vill. at Lakeridge, LLC , --- U.S. ----, 138 S. Ct. 960 , 967, 200 L.Ed.2d 218 (2018). We review a district court's statutory interpretation de novo. Bartling v. Fruehauf Corp. , 29 F.3d 1062 , 1069 (6th Cir. 1994).

The parties here present two mixed questions of law and fact: (1) whether the Safelite Plan is an employee pension benefit plan covered by 29 U.S.C. § 1002 (2)(A)(ii), and (2) whether the Safelite Plan is exempt from ERISA pursuant to DOL regulation 29 C.F.R. § 2510.3-2 (c). The facts of the case are undisputed, and these mixed questions "entail[ ] primarily legal ... work" in interpreting the language of the statute and the bonus plan regulation. U.S. Bank , 138 S. Ct. at 967 . We therefore review de novo the district court's interpretation of ERISA and the bonus plan regulation.

A. ERISA Employee Benefit Plans

The parties do not dispute that the Safelite Plan constitutes a "plan" for purposes of ERISA. See Hughes v. Zurz , 298 F. App'x 404 , 412 (6th Cir. 2008) (citing Donovan v. Dillingham , 688 F.2d 1367 , 1372 (11th Cir. 1982) ). The question then is whether the Safelite Plan is an ERISA employee pension benefit plan. See Buchanan v. General Motors, LLC , 597 F. App'x 305 , 309 (6th Cir. 2015). Such a plan must satisfy the statutory requirements that "by its express terms or as a result of surrounding circumstances," the plan either:

(i) provides retirement income to employees, or

(ii) results in a deferral of income by employees for periods extending to the termination of covered employment or beyond ....

29 U.S.C. § 1002 (2)(A). The parties agree that the Safelite Plan does not fall under subsection (i) but disagree about the application of subsection (ii). We turn now to § 1002(2)(A)(ii).

1. Statutory Language

The starting point is the language of the statute. See Hale v. Johnson , 845 F.3d 224 , 227 (6th Cir. 2016). "Where the statute's language is clear and unambiguous and the statutory framework is coherent and consistent, 'the sole function of the courts is to enforce it according to its terms.' " Id. (quoting United States v. Ron Pair Enters., Inc. , 489 U.S. 235 , 241, 109 S.Ct. 1026 , 103 L.Ed.2d 290 (1989) ). But "we must take care not to interpret the language [of a statute] in a vacuum;

*434 instead, we must look to the 'structure, history, and purpose' of the statutory scheme." Id. (quoting Abramski v. United States , 573 U.S. 169 , 179, 134 S.Ct. 2259 , 189 L.Ed.2d 262 (2014) ).

ERISA is a complex statute, but its purpose is simple: to establish a "uniform regulatory regime" for plan administration that protects monies belonging to plan beneficiaries while such funds are held and managed by others. Milby v. MCMC LLC , 844 F.3d 605 , 609 (6th Cir. 2016) (quoting Aetna Health Inc. v. Davila , 542 U.S. 200 , 208, 124 S.Ct. 2488 , 159 L.Ed.2d 312 (2004) ). In creating ERISA, Congress intended "to protect interstate commerce and the interests of participants in employee benefit plans and their beneficiaries" by (1) "requiring the disclosure and reporting to participants and beneficiaries"; (2) "establishing standards of conduct, responsibility, and obligation for fiduciaries of employee benefit plans"; and (3) "providing for appropriate remedies, sanctions, and ready access to the Federal courts." 29 U.S.C. § 1001 (b). As we have previously noted, "ERISA's purpose is among the broadest, if not the broadest, recognized by the Supreme Court," Sherfel v. Newson , 768 F.3d 561 , 564 (6th Cir. 2014), and Congress purposefully designed the scheme so that "[e]mployers can establish ERISA plans rather easily," Int'l Resources, Inc. v. New York Life Ins. Co. , 950 F.2d 294 , 297 (6th Cir. 1991) (citation and internal quotation marks omitted). This integrated system envisions that "any state-law cause of action that duplicates, supplements, or supplants the ERISA civil enforcement remedy conflicts with the clear congressional intent to make the ERISA remedy exclusive and is therefore pre-empted." Hogan v. Jacobson , 823 F.3d 872 , 879 (6th Cir. 2016) (quoting Davila , 542 U.S. at 209 , 124 S.Ct. 2488 ). At issue here is whether Wilson's state law claims are preempted because the Safelite Plan is an employee pension benefit plan covered by ERISA and not exempted as a bonus plan.

Keeping this statutory framework and purpose in mind, we examine the statutory definition. An "employee pension benefit plan" is established where a "plan, fund, or program ... by its express terms or as a result of surrounding circumstances ... results in a deferral of income by employees for periods extending to the termination of covered employment or beyond." 29 U.S.C. § 1002 (2)(A)(ii).

We begin with the meaning of "results," a key word in subsection (ii). The ordinary meaning of the word "results," as specified by the Supreme Court, is that "[a] thing 'results' when it '[a]rise[s] as an effect, issue, or outcome from some action, process or design.' " Burrage v. United States , 571 U.S. 204 , 210-11, 134 S.Ct. 881 , 187 L.Ed.2d 715 (2014) (quoting 2 The New Shorter Oxford English Dictionary 2570 (1993)). According to that ordinary meaning, a plan is an employee pension benefit plan when a "deferral of income by employees," 29 U.S.C. § 1002 (2)(A)(ii), arises as an "effect, issue, or outcome from" the provisions of that plan. Burrage , 571 U.S. at 210 , 134 S.Ct. 881 . Wilson posits that to fit within the initial phrase of subsection (ii), "results in a deferral of income," a plan "must require" deferrals to the termination of covered employment or beyond. But "results in" and "requires" are not synonymous. Moreover, Congress specifically used the words "require" or "requirement" in numerous other sections of Title I of ERISA but instead chose "results" here. See, e.g. , 29 U.S.C. § 1107 (b)(2)(B)(i) ("This paragraph shall apply to any eligible individual account plan if any portion of the plan's applicable elective deferrals (or earnings allocable thereto) are required to be invested in *435 qualifying employer securities or qualifying employer real property or both. ...") (emphasis added). "[W]here Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion." N. Fork Coal Corp. v. Fed. Mine Safety & Health Review Comm'n , 691 F.3d 735 , 741 (6th Cir. 2012) (quoting Russello v. United States , 464 U.S. 16 , 23, 104 S.Ct. 296 , 78 L.Ed.2d 17 (1983) ). In light of the ordinary meaning of the word "results" and Congress's exclusion of the word "requires," § 1002(2)(A)(ii) covers plans containing terms that have as an effect, issue, or outcome-even if not as a requirement-deferral of income by employees to periods extending to the termination of covered employment or beyond.

We turn next to the remainder of the statutory language: "a deferral of income by employees for periods extending to the termination of covered employment or beyond." 29 U.S.C. § 1002 (2)(A)(ii). Wilson reads the phrase to mean that employees must defer income "to the period that begins as one's employment ends," or "until termination," and argues that subsection (ii) does not cover plans that permit employees to withdraw deferred income before their termination from employment. Safelite argues that ERISA covers a plan that results in deferrals of income "for periods extending to the termination of covered employment or beyond" and also permits participants to receive in-service distributions of deferred income. Id . In essence, the question is whether a plan that allows for distributions both before and after termination can be an ERISA employee pension benefit plan.

Once again, we examine the statutory language, recalling "the cardinal principle of statutory construction that it is [the Court's] duty to give effect, if possible, to every clause and word of a statute rather than to emasculate an entire section." Bennett v. Spear , 520 U.S. 154 , 173, 117 S.Ct. 1154 , 137 L.Ed.2d 281 (1997) (quotation marks and alterations omitted). Subsection (ii) does not specify deferral of income "until termination" or "to termination"; rather, it says "for periods extending to the termination." Thus, deferrals may occur for various "periods," and those periods may last up to and/or beyond termination. Subsection (ii) covers a wide array of plans and does not exclude plans that give participants the option to receive in-service distributions. See Modzelewski v. Resolution Tr. Corp. , 14 F.3d 1374 , 1377 (9th Cir. 1994) ("Because ERISA's definition of a pension plan is so broad, virtually any contract that provides for some type of deferred compensation will also establish a de facto pension plan. ...").

2. DOL Exemption for Bonus Plans

The DOL promulgated regulations that "clarif[y] the limits" of the term "employee pension benefit plan" for purposes of Title I of ERISA. 29 C.F.R. § 2510.3-2 (a). By regulation, employee pension benefit plans do not include "payments made by an employer to some or all of its employees as bonuses for work performed, unless such payments are systematically deferred to the termination of covered employment or beyond, or so as to provide retirement income to employees." 29 C.F.R. § 2510.3-2 (c).

This regulation envisions bonuses, not pay for regular compensation, such as annual salaries. A "classic 'bonus' situation" involves "reward (higher cash value) for superior performance (higher corporate earnings)." Emmenegger v. Bull Moose Tube Co. , 197 F.3d 929 , 933 (8th Cir. 1999). Generally, a bonus plan's terms state that the plan's express purpose is to pay a *436 financial "bonus" or "additional incentive" to employees to encourage performance or retention. See, e.g. , id. ; Oatway v. Am. Int'l Grp., Inc. , 325 F.3d 184 , 189 (3d Cir. 2003) ; Hagel v. United Land Co. , 759 F. Supp. 1199 , 1200 (E.D. Va. 1991).

A bonus plan may defer payment of bonuses and remain exempt, "unless such payments are systematically deferred to the termination of covered employment or beyond, or so as to provide retirement income to employees." 29 C.F.R. § 2510.3-2 (c) (emphasis added). Post-termination payments of bonuses may result "incidental[ly]," though not systematically, and the plan would still be exempted from ERISA coverage. Oatway , 325 F.3d at 189 (finding no systematic deferral where the plan granted employees stock options that could be exercised on a yearly basis and would expire if not exercised in ten years); see also Murphy v. Inexco Oil Co. , 611 F.2d 570 , 575 (5th Cir. 1980) (finding no systematic deferral where the plan rewarded employees with royalty interests in a business prospect and paid the royalties annually). A bonus plan may also provide for certain in-service distributions according to a fixed vesting schedule where a deferral of bonus payments to termination "would only occur by happenstance." Emmenegger , 197 F.3d at 931-34 ; see also McKinsey v. Sentry Ins. , 986 F.2d 401 , 406 (10th Cir. 1993) ; Killian v. McCulloch , 850 F. Supp. 1239 , 1244 (E.D. Pa. 1994).

3. ERISA Coverage

In summary, ERISA § 1002(2)(A)(ii) covers any plan that, "by its express terms or as a result of surrounding circumstances," results in a deferral of income by participants for periods extending to the end of the participant's employment or beyond. The statutory language does not categorically exclude all plans that contain options to receive in-service distributions or to make other deferral elections. 29 U.S.C. § 1002 (2)(A)(ii). Regulations exclude from the definition of employee pension benefit plans a bonus plan providing for in-service distributions that are not systematically deferred to termination or for retirement income purposes. 29 C.F.R. § 2510.3-2 (c). We now apply this statutory and regulatory framework to the Safelite Plan.

B. The Safelite Plan

We begin by examining the Plan's express terms, looking first to the statement of purpose specified in the plan document. The first section of the Safelite Plan, titled "Purpose," states that "[t]he Company adopts this Plan ... to provide certain benefits to Eligible Employees" as a "program of deferred compensation within the meaning of Title I of ERISA." The express terms of the Safelite Plan indicate that the Plan means to provide the benefit of deferred compensation and intends to be covered by ERISA. See Rich v. Shrader , 823 F.3d 1205 , 1210 (9th Cir. 2016) (holding that "whether the primary purpose of the plan is to provide deferred compensation or other retirement benefits" is a "paramount consideration" in determining whether a plan falls under § 1002(2)(A)(i) or (ii) ).

Wilson does not dispute that the Safelite Plan is expressly established to be a deferred compensation plan. Wilson disputes whether the Plan "results in a deferral of income by employees for periods extending to the termination of covered employment or beyond." 29 U.S.C. § 1002 (2)(A)(ii). The Safelite Plan contains a default distribution provision specifying that deferred income will be distributed after a participant's termination from employment. Relevant sections state:

(a) Default Distribution. With respect to each Sub-Account of a Member, unless *437 a different election is made under Section 5.01(b), such Sub-Account will be paid in a lump sum as soon as administratively feasible after the Member Terminates, but in no case later than 60 days after the date the Member Terminates.

(b) Distribution Election . Notwithstanding the foregoing, with respect to each Sub-Account of a Member:

(i) The Member may elect to receive (or begin to receive) a distribution of the Sub-Account as soon as administratively feasible, but no later than 30 days, following (A) January 1 of the calendar year designated by the Member on a properly submitted Election Form or (B) his or her Disability; and

(ii) The Member may elect to receive the distribution described in Section 5.01(b)(i) in a lump sum payment or monthly (for distributions other than with respect to Disability) or annual payments over a period of up to ten years as designated by the Member on a properly submitted Election Form ....

In order to participate in the Plan, each eligible individual was required to submit an election form specifying the dollar amount to be deferred not only from his TIP Amount, but also from his yearly compensation and from his bonuses. The election form provided spaces for a participant to indicate if he wanted the deferred income distributed in a lump sum, in monthly payments, or in yearly payments. If those spaces were not filled, the default distribution was in "a lump sum as soon as administratively feasible after the Member Terminates."

The Fifth Circuit has addressed an income deferral plan similar to Safelite's. See Tolbert v. RBC Capital Mkts. Corp. , 758 F.3d 619 , 625-26 (5th Cir. 2014). In Tolbert , plan participants could defer compensation to be earned in the upcoming year and elect to have their accounts distributed either before or after termination of employment. Id. at 622-23 . The default distribution date was the vesting date, and vesting occurred automatically upon separation from employment. Id. at 622 . The Fifth Circuit held that because "[t]he 'express terms' of the [plan] also contemplate employees deferring income 'to the termination of covered employment or beyond[,]' .... the [plan] fits comfortably within the meaning of subsection (ii)." Id. at 626 .

As in Tolbert , at issue here is whether the Safelite Plan falls within the deferral of income provision of § 1002(2)(A)(ii), not the retirement income provision of subsection (i). The Safelite Plan expressly provides for employees to defer income from several sources to the future and authorizes options for payment of deferred income both before and after termination. Also comparable to Tolbert , the Plan's default provision provides for distributions to occur after termination of employment. As the district court correctly noted, "[t]he statute does not mandate that 'all deferrals extend to the termination of employment' or that payments be 'systematically deferred' until termination." Thus, the Safelite Plan contemplates that deferral occurs "for periods extending to the termination of covered employment or beyond" and fits within the meaning of subsection (ii).

The "surrounding circumstances," see 29. U.S.C. § 1002(2)(A), moreover, also point to a finding that the Safelite Plan falls within subsection (ii). In his complaint, Wilson alleges both that "[t]he Safelite Plan was designed to allow [him] to defer a significant portion of his income to later years for better tax treatment," and that he did defer income for years after he ceased to work for Safelite in 2008. Another Safelite executive declared that in December *438 2006, Wilson had urged the Board to adopt a deferred compensation plan to permit the eligible executives "to defer all or substantially all of their bonus, performance, and TIP payments until after they had left Safelite's employ." The record demonstrates that this is how the Plan was designed to work and was consistently administered-participants deferred income for various "periods," including periods extending up to and/or beyond termination.

To protect the interests of the participants and beneficiaries, Congress chose to subject certain plans that govern deferred income of employees to a uniform regulatory system and comprehensive civil enforcement scheme. To determine if the Safelite Plan is covered, we look to its design and administration, applying the language of the statute to the Plan's express terms and/or its surrounding circumstances. That decision is not dependent on the behavior of individual plan participants. As shown by the provisions of its plan documents and evidence in the record, the Safelite Plan provided for deferral of income within the statutory specifications of ERISA. The Plan "results in a deferral of income by employees for periods extending to the termination of covered employment or beyond." It is an ERISA employee pension benefit plan pursuant to § 1002(2)(A)(ii).

Wilson also argues that the Safelite Plan falls under the bonus plan exemption from ERISA coverage, 29 C.F.R. § 2510.3-2 (c). But unlike other plans found to be bonus plans, the Safelite Plan does not state an intention to provide financial incentives for employee performance or retention and does not explicitly operate as a bonus plan. The Plan also permits deferral of several types of employee income: participants' base salaries, annual bonuses, and the TIP Amount. Because the Safelite Plan is not designed as a bonus plan and instead distributes deferred amounts of non-bonus income, it is not a plan providing for payments made "as bonuses for work performed." 29 C.F.R. § 2510.3-2 (c).

Wilson asserts that because the Safelite Plan defers and ultimately distributes some incentive payments triggered under the terms of the TIP, the Plan's distributions are themselves made "as bonuses." But the cases on which he relies involved single plans which, by each plan's express terms, paid employees bonuses or provided stock options according to a specific vesting schedule. See McKinsey , 986 F.2d at 406 ; Lafian v. Elec. Data Sys. Corp. , 856 F. Supp. 339 , 345 (E.D. Mich. 1994) ; Copley v. Medpace, Inc. , 1:12-CV-426, 2013 WL 589158 , at *6 (S.D. Ohio Feb. 14, 2013), report and recommendation adopted , 12CV426, 2013 WL 1196609 (S.D. Ohio Mar. 25, 2013). As noted by the district court, Wilson's argument conflates the Safelite Plan, upon which the complaint rests, with the TIP, which is not challenged in this case.

We therefore agree with the district court that 29 C.F.R. § 2510.3-2 (c) does not support finding that "deferring bonuses provided for under a separate agreement transforms a deferred compensation plan into a bonus plan." The Safelite Plan is not a bonus plan and is not exempted from ERISA coverage.

III. CONCLUSION

Because the Safelite Plan qualifies as an employee pension benefit plan under 29 U.S.C. § 1002 (2)(A)(ii) and is not a bonus plan as defined in 29 C.F.R. § 2510.3-2 (c), we AFFIRM the district court's dismissal of the complaint.

CONCURRING IN PART AND IN THE JUDGMENT

THAPAR, Circuit Judge, concurring in part and concurring in the judgment.

I agree with the majority's textual analysis and concur to the extent the majority *439 rests on that analysis. Because the text of 29 U.S.C. § 1002 (2)(A)(ii) is clear, we should go no further. And the text is clear, as many tried-and-true tools of interpretation confirm. But so does one more: corpus linguistics. Courts should consider adding this tool to their belts.

I.

This case, like many others, asks us to interpret the text of a statute. Faced with the parties' conflicting understandings, we must figure out the most reasonable interpretation. When doing so, our job is not to replace the text that Congress enacted or explore the alleged purposes behind that statute. See Henry Schein, Inc. v. Archer & White Sales, Inc. , --- U.S. ----, 139 S. Ct. 524 , 530, 202 L.Ed.2d 480 (2019) ("Congress designed the Act in a specific way, and it is not our proper role to redesign the statute."). In fact, the subjective intent of the elected officials who enacted the statute is irrelevant. Our elected officials have many different reasons why they pass a certain statute or incorporate a specific provision. And we simply cannot know which of those varied reasons should control.

Instead, we operate under the fundamental principle that our elected officials write laws that all of us can understand by simply reading them. United States v. Davis , --- U.S. ----, 139 S.Ct. 2319 , 2323, --- L.Ed.2d ---- (2019) ("In our constitutional order, a vague law is no law at all."). This principle is baked into our constitutional design. For one, it gives meaning to having a republic. As James Madison argued, it makes little sense to elect people to govern if those people pass laws that are "so incoherent that they cannot be understood." See The Federalist No. 62, at 421 (James Madison) (J. Cooke ed., 1961). Such laws would "leave people no sure way to know what consequences will attach to their conduct," Davis , 139 S.Ct. at 2323, and would thus favor the "moneyed few over the industrious and uniformed mass of the people." The Federalist No. 62, supra , at 421. For another, the separation-of-powers principle that preserves our republic dictates that elected officials write the laws-not judges. Davis , 139 S.Ct. at 2323 ("Only the people's elected representatives in Congress have the power to write new federal ... laws."). If judges can simply rewrite statutes, "we would risk amending legislation outside the 'single, finely wrought and exhaustively considered, procedure' the Constitution commands." New Prime Inc. v. Oliveira , --- U.S. ----, 139 S. Ct. 532 , 539, 202 L.Ed.2d 536 (2019) (quoting INS v. Chadha , 462 U.S. 919 , 951, 103 S.Ct. 2764 , 77 L.Ed.2d 317 (1983) ). Thus, to fulfill these principles, we interpret laws with their "ordinary meaning at the time Congress enacted the statute." Id. (internal alterations, quotation marks, and citation omitted). Or, to put it another way, we look at how an ordinary person would normally understand the words that Congress used given the circumstances in which Congress used them. Only then can we give "ordinary people fair warning about what the law demands of them." Davis , 139 S.Ct. at 2323.

But words often have multiple permissible meanings. And parties will dispute which of a word's permissible meanings does, in fact, prove to be its ordinary one, given the statutory context. See Thomas R. Lee & Stephen C. Mouritsen, Judging Ordinary Meaning , 127 Yale L.J. 788 , 800, 802 (2018) (describing how words have possible, common, most frequent, exclusive, and prototypical meanings).

To assist in this sometimes-difficult task, judges and lawyers can utilize a variety of tools. Brett M. Kavanaugh, Fixing Statutory Interpretation , 129 Harv. L. Rev. 2118 , 2145 (2016) (book review); see generally *440 Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts (2012). For example, judges routinely consult the canons of interpretation, especially those which help us "understand the English language." Kavanaugh, supra , at 2145. We also look to other statutes or the pre-existing common law for context or to better understand a term's meaning. See Hall v. Hall , --- U.S. ----, 138 S. Ct. 1118 , 1128, 200 L.Ed.2d 399 (2018) ("[I]f a word is obviously transplanted from another legal source, whether the common law or other legislation, it brings the old soil with it." (quoting Felix Frankfurter, Some Reflections on the Reading of Statutes , 47 Colum. L. Rev. 527 , 537 (1947))); see also Scalia & Garner, supra , §§ 25, 53-54. And we need look no further than the majority opinion to see the value that dictionaries bring to our interpretive endeavor.

We ought to embrace another tool to ascertain the ordinary meaning of the words in a statute. This tool-corpus linguistics-draws on the common knowledge of the lay person by showing us the ordinary uses of words in our common language. How does it work? Corpus linguistics allows lawyers to use a searchable database to find specific examples of how a word was used at any given time. State v. Rasabout , 356 P.3d 1258 , 1275-76, 1289 (Utah 2015) (Lee, A.C.J., concurring in part and concurring in the judgment). These databases, available mostly online, contain millions of examples of everyday word usage (taken from spoken words, works of fiction, magazines, newspapers, and academic works). See, e.g. , Corpus of Contemporary American English , BYU, https://corpus.byu.edu/coca/help/texts.asp (listing types of sources); Corpus of Historical American English , BYU, http://www.english-corpora.org/coha/. Lawyers can search these databases for the ordinary meaning of statutory language like "results in." The corresponding search results will yield a broader and more empirically-based understanding of the ordinary meaning of a word or phrase by giving us different situations in which the word or phrase was used across a wide variety of common usages. See Lee & Mouritsen, supra , at 831 ("Linguistic corpora can perform a variety of tasks that cannot be performed by human linguistic intuition alone."). In short, corpus linguistics is a powerful tool for discerning how the public would have understood a statute's text at the time it was enacted.

Of course, corpus linguistics is one tool-new to lawyers and continuing to develop-but not the whole toolbox. Its foremost value may come in those difficult cases where statutes split and dictionaries diverge. In those cases, corpus linguistics can serve as a cross-check on established methods of interpretation (and vice versa). See Lawrence B. Solum, Triangulating Public Meaning: Corpus Linguistics, Immersion, and the Constitutional Record , 2017 BYU L. Rev. 1621 , 1669-70 (2017) ("[C]orpus linguistics allows for rigorous intersubjective validation of individual subjective judgments about word meaning."); see also Clark D. Cunningham et al., Plain Meaning and Hard Cases , 103 Yale L.J. 1561 , 1566 (1994) (book review) (arguing that empirical methods may assist judges in hard cases of ordinary meaning). This cross-check can provide both judges and parties with greater certainty about the meaning of words in a statute.

II.

The other concurring opinion argues that we should not add corpus linguistics to the judicial toolkit for several reasons. The first is methodological-corpora are not representative because of their sources. For instance, a corpus search for "flood" may lead to an overinclusion of *441 newspaper articles talking about giant flood waters rather than basements flooding. But the entire practice of law-and certainly the practice of interpretation-involves judgment calls about whether a particular source is relevant. And, at least with corpus linguistics, those calls can be vetted by the public in a more transparent way. Cf. Muscarello v. United States , 524 U.S. 125 , 142-43, 118 S.Ct. 1911 , 141 L.Ed.2d 111 (1998) (Ginsburg, J., dissenting) (criticizing the majority opinion for selective and non-transparent examples of word use). That is more than can be said of the alternative, which, as Justice Lee has thoughtfully noted, is for a judge to use his or her intuition-something far less representative and frankly far less "democratic." See Rasabout , 356 P.3d at 1274-75 (Lee, A.C.J., concurring in part and concurring in the judgment). Plus, the danger of judges relying upon their own intuition is that we introduce other risks, like confirmation bias. Id. at 1274 . Judges may unintentionally give greater weight to those definitions that match up with their preconceived notions of a word's meaning. We cannot get away from confirmation bias altogether, but we can surely check our intuition against additional sources of a word's meaning. The corpus allows us to do this.

Second, the other concurring opinion argues that the use of corpus linguistics will descend into mere rote frequency analysis; judges will simply pick the use of the word that shows up the most. Yet judges who use corpora do not become automatons of algorithms. They will still need to exercise judgment consistent with the use of the other tools of statutory interpretation. Sometimes the most frequent use of a word will line up with its ordinary meaning as used in a statute. Sometimes it will not. The data from the corpus will provide a helpful set of information in making that interpretive decision. But the judge must make the ultimate decision after considering multiple tools.

Third, the other concurring opinion suggests that corpus linguistics is redundant when compared with another tool-dictionaries. Expert lexicographers already do corpus linguistics when compiling dictionaries, so, the argument goes, when judges use corpus linguistics, they become unnecessary and unhelpful armchair lexicographers. But the use of corpus linguistics improves upon dictionaries by helping pinpoint the ordinary uses of a word at the time a statute was enacted. For example, when a court considers a dictionary definition, it looks at a dictionary from that time period. See New Prime , 139 S. Ct. at 539-40 & n.1. But the usage examples in those dictionaries often come from a time before the dictionary was published. See Lee & Mouritsen, supra , at 808-09; Henry M. Hart, Jr. & Albert M. Sacks, The Legal Process: Basic Problems in the Making and Application of Law 1190 (William N. Eskridge, Jr. & Phillip P. Frickey eds., 1994) ("An unabridged dictionary is simply a[ ] historical record, not necessarily all-inclusive, of the meanings which words in fact have borne, in the judgment of the editors, in the writings of reputable authors."); see also id. at 1375-76. So the dictionary definition may actually tell us the ordinary meaning at a time long before Congress enacted the statute. See Lee & Mouritsen, supra , at 809; Scalia & Garner, supra , Appendix A at 419 (noting that dictionaries lag behind the times). And in many cases (like the ones discussed below), both the majority and dissent will point to dictionaries without any clear resolution. Instead of relying on just a few sample sentences in the dictionary, the corpus develops a broader picture of how words were actually used when Congress passed the statute.

*442 Plus, "[w]e judges are experts on one thing-interpreting the law." Rasabout , 356 P.3d at 1285 (Lee, A.C.J., concurring in part and concurring in the judgment) (emphasis omitted). Corpus linguistics is just one variation on a very old theme in this field of expertise. Judges often consider the context of words-both within and beyond the text of the statute in dispute. See Scalia & Garner, supra , § 31 (detailing the "noscitur a sociis" canon of interpretation); see also Caleb Nelson, What is Textualism? , 91 Va. L. Rev. 347 , 355 (2005). Judges look to contemporaneous judicial decisions. See, e.g. , New Prime , 139 S. Ct. at 540. They look to seemingly common phrases. See, e.g. , FCC v. AT&T Inc. , 562 U.S. 397 , 403-04, 131 S.Ct. 1177 , 179 L.Ed.2d 132 (2011) (considering how the word "personal" is used in "personal life" and "personal expenses"). And, for constitutional cases, they look to word use in the Anti-Federalist and Federalist Papers. See, e.g. , United States v. Lopez , 514 U.S. 549 , 586, 115 S.Ct. 1624 , 131 L.Ed.2d 626 (1995) (Thomas, J., concurring). While sometimes this "enterprise may implicate disciplines or fields of study on which we lack expertise, [it] is no reason to raise the white flag" and forgo considering corpus linguistics. Rasabout , 356 P.3d at 1285 (Lee, A.C.J., concurring in part and concurring in the judgment). Instead, judges should do what they have always done-"summon all our faculties as best we can" and take advantage of adversarial briefing. See id.

In sum, I agree that corpus linguistics is not the only tool we should use, but it is an important tool that can assist us in figuring out the meaning of a term.

III.

Some examples help prove the point. Consider Smith v. United States , 508 U.S. 223 , 113 S.Ct. 2050 , 124 L.Ed.2d 138 (1993). In that case, the majority and dissent debated whether exchanging a firearm for drugs constitutes "use" of a firearm "during and in relation to a ... drug trafficking crime." Id. at 228 , 113 S.Ct. 2050 (quoting 18 U.S.C. § 924 (c)(1) ) (internal quotation marks omitted); id. at 241-43 , 113 S.Ct. 2050 (Scalia, J., dissenting). The Smith majority noted that, because "use" was not defined within the statute, the Court needed to determine "its ordinary or natural meaning." Id. at 228 , 113 S.Ct. 2050 . Thus, the majority consulted an 1884 case and two dictionaries (Webster's and Black's) before deciding that "use" meant "to employ" or "to derive service from." Id. at 229 , 113 S.Ct. 2050 (citations omitted). And, in trading a firearm for drugs, the majority concluded that the defendant " 'used' or 'employed' " that firearm. Id. Three justices disagreed. Id. at 241-42 , 113 S.Ct. 2050 (Scalia, J., dissenting). Writing for these dissenters, Justice Scalia first noted that, when we "use an instrumentality," we ordinarily mean that we will "use it for its intended purpose." Id. at 242 , 113 S.Ct. 2050 . He colorfully pointed out that when someone asks, " 'Do you use a cane?,' he is not inquiring whether you have your grandfather's silver-handled walking stick on display in the hall; he wants to know whether you walk with a cane." Id. So when someone says, "Do not use a firearm," the natural reading of that phrase is that the listener should not use a firearm as a weapon -not that the listener cannot scratch his head with the firearm. Id. at 242-43 , 113 S.Ct. 2050 .

Could corpus linguistics have helped in this debate? I think so. In a study of the Corpus of Contemporary American English, the authors found 159 instances where the verb "use" was followed by a noun representing a weapon (like "gun" or "rifle"). Stefan Th. Gries & Brian G. Slocum, *443 Ordinary Meaning and Corpus Linguistics , 2017 BYU L. Rev. 1417 , 1459 (2017). Of these 159 instances, there was not a single time that "use" signified a barter. Id. at 1460 (noting further that 88% were clearly not barter and many involved discharging or brandishing, while 12% were probably not barter when coded "in the most conservative way," meaning the "comprehender would have to bend over backwards semantically" to demonstrate that "use" meant "barter"). And, to ensure that this conclusion was not simply based on the most frequent use of the word, the authors of the study ensured that these examples were sufficiently dispersed in the corpus, i.e., that they did not come only from one source or one type of source (like legal writings). Id. at 1459-60 . It is hard to say this evidence would have been neither compelling nor relevant to the debate between the majority and dissent in Smith .

Take another quick example. In Muscarello , the Supreme Court debated whether carrying a firearm meant carrying it on one's person or carrying it more generally (as in a car). 524 U.S. at 126-27 , 118 S.Ct. 1911 . The majority concluded that a person who transported a firearm in a locked glove compartment or in the trunk of his car was guilty of "carry[ing] a firearm during and in relation to a drug trafficking crime." Id. (quoting 18 U.S.C. § 924 (c)(1) ) (internal quotation marks omitted). In contrast, Justice Ginsburg, writing for four dissenting justices, said that the term "carr[ying]" contextually meant to carry on one's person. Id. at 139-40 , 118 S.Ct. 1911 (Ginsburg, J., dissenting). The debate between the majority and dissent is instructive, as the justices themselves performed ad hoc corpus analyses. The majority looked at how "carry" was used in the King James Bible, Robinson Crusoe, and Moby Dick, in addition to "survey[ing] modern press usage, albeit crudely, by searching computerized newspaper databases." Id. at 129 , 118 S.Ct. 1911 (citations omitted). Justice Ginsburg offered another ad hoc corpus search with her own examples. Id. at 143-44 , 118 S.Ct. 1911 (Ginsburg, J., dissenting).

Like with Smith , an actual corpus analysis has since been performed. See Lee & Mouritsen, supra , at 845, 847. The evidence suggests that, generally (but not always), people use "carry" when referring to someone personally carrying an object. Id. at 847; see also Neal Goldfarb, A Lawyer's Introduction to Meaning in the Framework of Corpus Linguistics , 2017 BYU L. Rev. 1359 , 1405-12 (2017) (using a different corpus but reaching a similar conclusion). Again, it is difficult to say the corpus analysis would not have assisted the lawyers and the Court.

IV.

Now for the case at bar. Dan Wilson sued Safelite Group, Inc. ("Safelite") under Ohio law about its Nonqualified Deferred Compensation Plan ("Safelite Plan"). We have grappled with whether ERISA (a 1974 statute) applies to the Safelite Plan. If so, Wilson's state law claims are preempted, and his lawsuit is over. As relevant here, ERISA covers a company's deferred income plan when the plan "results in a deferral of income by employees for periods extending to the termination of covered employment or beyond." 29 U.S.C. § 1002 (2)(A)(ii). Unsurprisingly, Wilson argues that his Safelite Plan does not fall under ERISA-principally because the plan allowed for distributions while he was still employed. He reads the statute to mean that ERISA covers only those plans that " require a deferral of income by employees for periods until the termination of covered employment or beyond." Because the plan allowed payments while employees still worked at Safelite, the plan *444 does not require all income to be deferred until termination. According to Wilson, that means ERISA does not apply.

In making his argument, Wilson claims that "results in" means "requires" and that "extending to" means essentially "until a certain time." "Results in," however, is a phrasal verb (verbs that end in a preposition or participle) that, at the time ERISA was passed in 1974, meant "to cause" or "have (something) as a result." Longman Dictionary of Phrasal Verbs 506 (1983); see also Scalia & Garner, supra , Appendix A at 419 (noting that to understand the meaning of a word in a statute, it is "quite permissible" to consult a dictionary published a decade or so later because dictionaries "lag behind linguistic realities"). "Results in" does not mean "requires." And, as the majority explains, where Congress wanted to make something a requirement in the ERISA statute, it specifically did so by using the word "requires." See Majority Op. Part II.A.1. Similarly, when grappling with "extending to," the majority applied the surplusage canon to make sure we gave effect to the term "extending to" and correctly found that "extending to" does not preclude in-service distributions. "Extend to," another phrasal verb, meant "to reach" at the time ERISA was passed. Longman Dictionary of Phrasal Verbs 170 (1983). In sum, § 1002(2)(A)(ii) includes plans that "cause" the deferral of income by employees for periods "t[hat] reach" to the termination of covered employment or beyond. See Longman Dictionary of Phrasal Verbs 170, 506 (1983).

The corpus confirms this textual analysis. If we limit the searches to the 1960s and 1970s (the time period immediately before and during ERISA's enactment), we only get a few hundred results for each phrase in the Corpus of Historical American English. And there is simply no result-not one-where the phrasal verb "results in" could be read to mean "requires." To bring this to life, here are just a few of the many examples generated by the corpus search:

• "Two enterprises are competitive when an increase in the output of one results in a decrease in the output of the other." Emery N. Castle & Manning H. Becker, Farm Business Management: The Decision-Making Process (1962).

• "Army gets little gridiron help, because the cooler New York weather results in spring football running until late April." Federal Power Takes Over , Sports Illustrated (June 8, 1964).

• "The Federal Aviation Act frowns on price competition among the interstate airlines, and the CAB quickly pounces on any sign of rate-cutting. This results in competition being trivialized into 'booze wars,' 'lounge wars,' and fuselage decoration, with the passengers compelled to pay the bill." Peter H. Schuck, Why Regulation Fails , Harpers (September 1975).

• "Do not let an inexperienced guest attack [splitting logs] without instruction and supervised practice. Most of my guests tended to drive one wedge into a log as far as it would go and then find that the log had not split and they could not remove the wedge. This results in frustrated guests, a battered wedge head and unsplit wood." Patricia Crawford, Homesteading (1975).

For "extending to," the corpus search included "extend(s) to" as well. Here too, the analysis supports the majority's conclusion with only one example that arguably can be read to mean "until a certain point in time." Here are just a few, starting with the one that supports Wilson's reading:

*445 • "The hospital is being built by Hospital Corporation of America, which has a contract extending to 1980 to recruit employees and medical staff and to manage the hospital." Howard Dernton, Saudi Arabia , Saturday Evening Post (April 1974).

• "A new public plaza at Chambers Street will link the center of the island with the waterfront of the Hudson River, connecting with Battery Park City to the south and a shoreline park system extending to the north and south, the Mayor said." Charles G. Bennett, Downtown Renewal Plan Adds College for 5,000 , N.Y. Times (1968).

• "A favorite design is a heart-shape with two prongs extending to either side from the point of the heart." Alex W. Bealer, The Art of Blacksmithing (1969).

• "The knights support a worldwide program of medical aid and refugee relief that extends to 42 countries." Knightly Return , Time (June 21, 1968).

Ultimately, in this case, the corpus results serve as a method to check our work. And those results confirm that the majority's textual analysis is correct. In future cases where the ordinary meaning is debatable, like in Smith and Muscarello , the results could be determinative.

* * *

We seek ordinary meaning because laws are written for everyday people, and it is not our role as judges to rewrite those laws. I join the portions of the majority's opinion that faithfully apply traditional tools of textual analysis and that do not journey beyond the statutory text to discern ordinary meaning. In this case, a corpus linguistics analysis confirms the majority's analysis as well. In future cases, adversarial briefing on corpus linguistics can help courts as they roll up their sleeves and grapple with a term's ordinary meaning.

CONCURRENCE

JANE B. STRANCH, Circuit Judge, concurring.

I agree with Judge Thapar's concurrence that we need look no further when statutory text is clear. That is what the majority opinion does-first by recognizing ERISA's statutory framework and purpose as expressed by Congress in 29 U.S.C. § 1001 (b) and then by examining the statutory definition in § 1002(2)(A)(ii). I write separately to express some concerns about the concurrence's endorsement of "corpus linguistics," a proposed method of statutory interpretation described in a handful of recent state court opinions. 1 This tool invites judges to perform the same kind of subjective decision making that the concurrence otherwise cautions us to avoid. There are several reasons why we should decline this invitation.

The first is a practical problem. A keyword search using a corpus linguistics database will likely result in dozens, if not hundreds or thousands, of examples of a term's usage. How should courts make sense of all this information? First, we could count the number of times a term is used in the database (assuming appropriately selected parameters) and then decide that the most frequently used meaning is the ordinary meaning. But that approach would risk privileging the most newsworthy connotations of a term *446 over its ordinary meaning. See, e.g. , Carissa Byrne Hessick, Corpus Linguistics and the Criminal Law , 2017 B.Y.U. L. Rev. 1503, 1509 ("[A] corpus analysis may demonstrate that seventy percent of all mentions of the word 'flood' occur in the context of [ ] superstorms. But that does not tell us whether the average American would understand the statutory term 'flood' to include three inches of water in a homeowner's basement after a neighboring water main burst."). It would also fail to cull irrelevant results. If we use the database to determine the meaning of "results in" during the 1960s and 1970s (the era closest to when ERISA was drafted), we find examples of this term's usage in contexts that bear no relationship to our own. Does it matter, for purposes of deciding whether ERISA applies to Wilson's deferred compensation plan, how "results in" was used in a book about farm animal management in 1976, or in an article from Sports Illustrated about New York's cool spring weather in 1964? 2 I think it does not. And even if consulting this overinclusive data set might help judges "to avoid basing conclusions on a few speakers' idiosyncrasies," it is "the 'idiosyncrasies' of [Congress that] constitute the rule of law in this [country]. And the only way to identify those idiosyncrasies is through the text of the [U.S.] Code, which is wholly absent from [the corpus linguistics] data set." State v. Rasabout , 356 P.3d 1258 , 1266 (Utah 2015) (call numbers and internal quotation marks omitted). This suggests to me a disconnect between corpus linguistics and the judicial work of statutory interpretation.

Another approach would require the court to perform this culling process itself. For example, we could assume that the drafters employed popular, as opposed to technical or legal, language and decide which uses of "results in" during the 1960s and 1970s should be included in our analysis and which should not. But by what metric would we make that choice? Perhaps most could agree that a book about farm animal management is not relevant here. But what about an article reporting a union strike? Or one about federal tax penalties? 3 Such choices would require highly subjective, case-by-case determinations about the import and relevance of a given source. Textualists have long advised us to forgo that interpretive method. See, e.g. , Conroy v. Aniskoff , 507 U.S. 511 , 519, 113 S.Ct. 1562 , 123 L.Ed.2d 229 (1993) (Scalia, J., concurring in the judgment) ("The legislative history of [this] Act contains a variety of diverse personages, a selected few of whom-its 'friends'-the Court has introduced to us in support of its result. But there are many other faces in the crowd, most of which, I think, are set against today's result."). Legislative history tells us, at a minimum, how some of the statute's authors understood a term; corpus linguistics does not offer even that insight.

In part because of these practical problems, the use of corpus linguistics is a difficult and complex exercise. That is why, for centuries, we have left this task to the trained lexicographers who author the tool *447 we already employ-a dictionary. The concurrence describes much of what lexicographers do every day. See, e.g. , Phillip A. Rubin, War of the Words: How Courts Can Use Dictionaries Consistent with Textualist Principles , 60 Duke L.J. 167 , 179 (2010) ("Nowadays, most dictionary authors create a 'corpus,' which is a collection of whole or partial texts or recorded speech stored and indexed electronically so that individual words can be found quickly."); How Our Dictionaries Are Created , Oxford Dictionaries, https://www.oxforddictionaries.com/our-story/creating-dictionaries (last visited June 17, 2019) (noting that a dictionary's "team of expert lexicographers ... are constantly researching, analysing, and documenting languages as they change and develop").

Lexicographers engage in "research, but also decisionmaking: the primary job of the lexicographer in creating a dictionary is to determine meanings of words, and to determine what different meanings a word might have." Rubin, supra at 181. And because "[t]he line between one meaning and another is seldom clear," this process "leaves much of the final determination to the experienced judgment of the editorial staff." Id . The other concurrence argues that, notwithstanding their training, these experts sometimes select outdated or otherwise unreliable meanings for disputed terms. But I would not substitute the ad hoc selection process of individual judges for the "experienced judgment" of these trained scholars. Doing so would convert judges into armchair lexicographers, attempting the same work that dictionary authors have been performing for centuries. But unlike those experts, judges would shoulder this task without the specialized training necessary to make a reliable and neutral judgment call. Encouraging litigants to take on that same role would make the problem worse, not better.

I do not suggest that corpus linguistics can never assist judges in the difficult project of statutory interpretation. But, in the unusual case where such a tool might prove useful, I would leave this task to qualified experts, not to untrained judges and lawyers. See, e.g. , Brief for Professor Clark D. Cunningham, et al. as Amicus Curiae on Behalf of Neither Party, In Re: Donald J. Trump, President of the United States of America , No. 18-2486 (4th Cir. Jan. 29, 2019) (discussing use of corpus linguistics by professor of applied linguistics to help determine the meaning of "emoluments" during the founding era). And before we add corpus linguistics to our judicial toolkit, we should first remind ourselves what our toolkit is for. I agree with the concurrence that statutes ought to give "ordinary people fair warning about what the law demands of them." United States v. Davis , --- U.S. ----, 139 S. Ct. 2319 , 2323, --- L.Ed.2d ---- (2019). But that is the responsibility of legislators, not judges. Once the torch passes from Congress to the courts, our duty as judges is simply to determine what our elected members of Congress meant when they passed the statute-even if that is not the meaning we or the public might routinely employ. In most cases, adding corpus linguistics to our judicial toolkit would make it harder to focus on that narrow duty. This case underscores why. Our task here is to decide what Congress meant by "results in" and "extending to" when it defined the universe of employee pension benefit plans covered by ERISA. The other concurrence proposes that we divine that meaning not by considering ERISA's statutory framework or legislative history, but by looking to the language of an article from Sports Illustrated and a book about farm animal management. I struggle to see why those sources would tell us as much as, for example, what the legislature *448 told us about the structure and purpose of ERISA when it drafted the statute.

Underlying these practical usage issues is my concern with the implicit suggestion that corpus linguistics is a simple, objective tool capable of providing answers to the puzzle of statutory interpretation. The use of corpus linguistics brings us no closer to an objective method of statutory interpretation. Instead, it encourages judges to stray even further from our historic and common-sense considerations-including the "text, structure, history, and purpose" of a statute, Maracich v. Spears , 570 U.S. 48 , 76, 133 S.Ct. 2191 , 186 L.Ed.2d 275 (2013) (citation omitted)-that ought to guide our analysis.