MCI SALES AND SERVICE, INC., f/k/a Hausman Bus Sales, Inc. and Motor Coach Industries Mexico, S.A. de C.V., f/k/a Dina Autobuses, S.A. de C.V., Petitioners, v. James HINTON, Individually and as Representative of the Estate of Dolores Hinton, Deceased, et al., Respondents, 329 S.W.3d 475

Justice GUZMAN delivered the opinion of the Court, in which Justice HECHT, Justice WAINWRIGHT, Justice MEDINA, Justice JOHNSON, Justice WILLETT, and Justice LEHRMANN joined, and in which Chief Justice JEFFERSON joined as to Parts I and II. This appeal arises from a jury’s verdict in a suit brought against the manufacturer, importer, and distributor of a motorcoach. In 1995, when the motorcoach at issue here was manufactured, federal safety regulations governing the performance of these motorcoaches neither required nor forbade passenger seatbelts.

These same regulations allowed manufacturers to choose between several types of glazing materials for use in the motorcoaches’ windows, and a manufacturer complied by using one of the required types. We must decide whether that regulatory silence and that choice evidence a congressional intent to preempt a McLennan County jury’s finding that the manufacturer of a motor-coach should have installed passenger seatbelts and should have used another permitted type of glazing material.

Because we conclude that the jury’s verdict which is grounded in this state’s common law does not present any obstacle to the accomplishment of the federal regulatory scheme’s purpose, we hold that the federal safety standards at issue do not preempt state law. We also apply

Chapter 33 of the Texas Civil Practice and Remedies Code to a plan adopted by a bankruptcy court to apportion a debtor’s insurance proceeds among a group of creditors who filed claims against the bankruptcy estate. The unique plan allowed the claimants to either accept a mediated percentage of the proceeds or to litigate their claims before a special judge. Even if the claimants chose the latter course, them recovery was capped at 110% of the mediator’s award, and the claimants could agree at any time to full or partial distributions to any or all of the claimants.

We must decide if this plan renders the debtor — who purchased the insurance policy funding the plan and whose further liability was discharged in bankruptcy — a settling person under

Chapter 33 for purposes of determining proportionate liability. We conclude that it does. Accordingly, we affirm the court of appeals’ judgment and remand to the trial court for further proceedings consistent with this opinion. I. Background On February 14, 2003, a group of friends chartered a bus from Central Texas Trails to take them from Temple to Dallas for a concert. Heavy rain and fog reduced visibility, and as the bus crested a hill on Interstate 35 south of Waco, the driver saw that traffic had stopped due to an accident farther north.

He attempted to change lanes to increase his stopping distance, but another car cut him off, so he steered into the earthen median and lost control of the bus. It crossed the median into southbound traffic and collided with a large sport utility vehicle, spun counterclockwise, and tipped over on its right side. The bus slid across the southbound lanes and came to rest in the ditch on the far side of the road. Most of the large, non-laminated glass windows on the right side of the bus shattered when it tipped over. The passengers were tossed from their seats and some were ejected through the broken windows.

Five passengers were killed and several others were injured to varying degrees. The bus owner and operator, Central Texas Trails, Inc., Central Texas Bus Lines, Inc., and Kincannon Enterprises, Inc. (collectively Central Texas), filed for

Chapter 11 bankruptcy protection shortly after the accident. The bus crash victims filed creditor claims against Central Texas in the bankruptcy court. Central Texas maintained a $5 million liability insurance policy, and the carrier paid the policy limits into the bankruptcy court’s registry, creating a liability fund. The bus crash claimants participated in non-binding mediation, the goal of which was to formulate a plan for apportioning the fund.

The mediator assigned a percentage of the fund to each claimant, and these percentages were incorporated into a plan submitted to the bankruptcy court for approval, which was given on October 21, 2003. Under the “Apportionment Plan,” a claimant could accept the mediator’s percentage and immediately receive that portion of the liability fund.

If the claimant chose not to accept the mediator’s allocation, the claimant participated in a “Litigation Plan.” Under this plan, the claimants tried their claims to a special judge agreed to by the participants, and their recovery under this plan was capped at 110% of the mediator’s allocation. Further, the parties could agree at any time to approve a full or partial distribution to any or all participants. Central Texas’s tort liability in excess of the liability fund was discharged upon approval of its reorganization plan the following year.

On June 26, 2003, a group of the injured bus occupants (or their estates) and their relatives (the Plaintiffs) filed suit against Motor Coach Industries Mexico, S.A. de C.V., and MCI Sales and Service, Inc., the bus’s manufacturer, importer, and distributor (collectively MCI), alleging the motor-coach was defectively designed because it lacked passenger seatbelts and laminated-glass windows.

MCI attempted to join Central Texas and the bus driver as responsible third parties, but the trial court denied the motion and also refused to submit a question asking the jury if Central Texas and the driver were liable as responsible third parties or as settling parties to determine proportionate liability. Following a jury trial, the jury found for the Plaintiffs, making separate causative findings as to each claim. The jury found that the lack of seatbelts caused injuries to all of the Plaintiffs, and the lack of laminated-glass windows caused injuries to those ejected from the bus.

The jury awarded over $17 million in damages. After the jury’s verdict but before entry of judgment, the Plaintiffs, all of whom opted for resolution of their claims against Central Texas via the Litigation Plan, appeared before the special judge. The judge found that the bus driver’s negligence proximately caused the accident that produced the Plaintiffs’ injuries and that he was acting within the scope of his employment with Central Texas. Further, the judge determined that the jury’s damage awards in this case, with one exception, significantly exceeded the 110% maximum recovery under the Litigation Plan.

Limiting the one participant’s recovery to what the jury awarded, the judge capped the remaining damage awards at the 110% maximum. Because of the limited funds, however, the awards were prorated so that each participant received a relative percentage of the actual award. In the end, each Litigation Plan participant (again, with the one exception) received a sum that is within two percent of the amount allocated by the mediator. The bankruptcy judge approved the special judge’s report, and the payments were made.

Thereafter, the trial court in this case entered judgment, adjusting the damage awards to account for the funds received under the Litigation Plan. MCI appealed, and the court of appeals reversed and remanded. 272 S.W.3d 17, 20. As relevant here, the court rejected MCI’s preemption arguments but agreed that the trial court abused its discretion by not submitting a question to the jury regarding Central Texas and the bus driver’s proportionate liability as settling persons.

MCI then petitioned this Court for review of the preemption issues, and the Plaintiffs cross-petitioned for review of the proportionate responsibility issue. We granted both petitions and consider the issues in order, beginning with federal preemption. II. Federal Preemption The Supremacy Clause dictates that the “Constitution, and the Laws of the United States which shall be made in Pursuance thereof ... shall be the supreme Law of the Land; ... any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” U.S. CONST, art. VI, cl. 2.

Thus, a law passed by Congress — acting within its enumerated powers — and signed by the President preempts any state law to the contrary, rendering it without effect. Maryland v. Louisiana, 451 U.S. 725, 746, 101 S.Ct. 2114, 68 L.Ed.2d 576 (1981); Mills v. Warner Lambert Co., 157 S.W.3d 424, 426 (Tex.2005) (per curiam); see also Wyeth v.

Levine, — U.S. -, 129 S.Ct. 1187, 1206-07, 173 L.Ed.2d 51 (2009) (Thomas, J., concurring in the judgment) (noting the two structural limitations on the federal government’s power to preempt state laws, namely, the enumerated powers of Congress and the procedural requirements to enact a law, including bicameral passage and presentment to the President). Federal regulations properly adopted by an agency acting within its congressionally delegated authority likewise preempt a contrary state law. E.g., City of New York v. FCC, 486 U.S. 57, 63-64, 108 S.Ct. 1637, 100 L.Ed.2d 48 (1988) (citing La. Pub. Serv. Comm’n v.

FCC, 476 U.S. 355, 369, 106 S.Ct. 1890, 90 L.Ed.2d 369 (1986)). Congress may expressly preempt state law by means of statutory language, see, e.g., Cipollone v. Liggett Group, Inc., 505 U.S. 504, 516, 112 S.Ct. 2608, 120 L.Ed.2d 407 (1992); Great Dane Trailers, Inc. v. Estate of Wells, 52 S.W.3d 737, 743 (Tex.2001), or it may do so impliedly in one of two ways, by “so thoroughly occup[ying] a legislative field ‘as to make reasonable the inference that Congress left no room for the States to supplement it,’ ” Cipollone, 505 U.S. at 516, 112 S.Ct. 2608 (quoting Fidelity Fed. Sav. & Loan Ass’n v.

De la Cuesta, 458 U.S. 141, 153, 102 S.Ct. 3014, 73 L.Ed.2d 664 (1982) (quotation marks omitted)), or by enacting a law that actually conflicts with state law, see id. (citing Pac. Gas & Elec. Co. v. State Energy Res. Conservation & Dev. Comm’n, 461 U.S. 190, 204, 103 S.Ct. 1713, 75 L.Ed.2d 752 (1983)). A state law actually conflicts with a federal law when compliance with both is impossible or when the state law “stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.” Hines v.

Davidowitz, 312 U.S. 52, 67, 61 S.Ct. 399, 85 L.Ed. 581 (1941); accord Great Dane Trailers, 52 S.W.3d at 743. The latter kind of conflict preemption, sometimes called obstacle preemption, is the only theory of preemption advanced by MCI in this case. See Geier v. Am. Honda Motor Co., 529 U.S. 861, 868, 120 S.Ct. 1913, 146 L.Ed.2d 914 (2000) (holding that the National Traffic and Motor Vehicle Safety Act’s express preemption clause does not apply to common-law tort actions).

As the Supreme Court has observed, “the purpose of Congress is the ultimate touchstone in every pre-emption case.” Wyeth, 129 S.Ct. at 1194 (quotation marks omitted). When Congress has explicitly stated that its legislation preempts state law, that intent is plain. See, e.g., Geier, 529 U.S. at 867-68, 120 S.Ct. 1913 (concluding that the Safety Act’s express preemption clause preempts nonidentical standards contained in state legislation and regulations).

At other times, Congress does not expressly state its preemptive purpose, but such intent is discoverable through the statutory language and structure, as when Congress occupies a field of regulation and leaves no room for states to operate. See, e.g., City of Burbank v. Lockheed Air Terminal Inc., 411 U.S. 624, 633, 93 S.Ct. 1854, 36 L.Ed.2d 547 (1973) (“It is the pervasive nature of the scheme of federal regulation of aircraft noise that leads us to conclude that there is pre-emption.”).

But when preemption is premised on the obstacle a state law erects to accomplishing a federal purpose, divining that intent can be more challenging. To discover the federal purpose with which the state law is in conflict, we examine- the text of the relevant federal statute or regulation, the history of federal regulation in that subject area, and the agency’s statements construing the regulation. See Geier, 529 U.S. at 875-77, 120 S.Ct. 1913; O’Hara v. Gen.

Motors Corp., 508 F.3d 753, 759 (5th Cir.2007) (“To determine the federal policy expressed in [the regulation], this Court looks to the text of the regulation, the history of [the agency’s] regulation in this area, and [the agency’s] statements construing [the regulation].”). A “specific, formal agency statement identifying conflict” is not necessary to conclude that a conflict exists, Geier, 529 U.S. at 884, 120 S.Ct. 1913, and “[t]he weight we accord the agency’s explanation of state law’s impact on the federal scheme depends on its thoroughness, consistency, and persuasiveness,” Wyeth, 129 S.Ct. at 1201.

Courts must cautiously approach this interpretive task lest it become a “free wheeling judicial inquiry into whether a state [law] is in tension with federal objectives,” which “undercut[s] the principle that it is Congress rather than the courts that pre-empts state law.” Gade v. Nat’l Solid Wastes Mgmt. Ass’n, 505 U.S. 88, 111, 112 S.Ct. 2374, 120 L.Ed.2d 73 (1992) (Kennedy, J., concurring in part and concurring in the judgment). With this framework in mind, we turn to the relevant federal law that MCI asserts preempts the jury’s verdict. A.

The Federal Motor Vehicle Safety Standards Congress passed the National Traffic and Motor Vehicle Safety Act of 1966 (Safety Act or Act), to reduce traffic accidents and their resulting injuries. Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 33, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) (citing 15 U.S.C. § 1381 (1976) (current version at 49 U.S.C. § 30101)). Congress gave the Secretary of Transportation the authority to “prescribe motor vehicle safety standards.

Each standard shall be practicable, meet the need for motor vehicle safety, and be stated in objective terms.” 49 U.S.C. § 30111(a). A motor vehicle safety standard is “a minimum standard for motor vehicle or motor vehicle equipment performance.” Id. § 30102(a)(9). The Secretary of Transportation in turn delegated authority to create the safety standards to the Administrator of the National Highway Traffic Safety Administration (NHTSA). 49 C.F.R. § 1.50. Over the years, NHTSA has promulgated and modified the Federal Motor Vehicle Safety Standards (FMVSS), of which FMVSS 205 and 208 are at issue here.

See id. §§ 571.205, .208. 1. FMVSS 208 — Federal Regulation of Seatbelts In 1971, NHTSA issued FMVSS 208, which governed occupant crash protection and mandated seatbelts in motorcoaches for the driver only. See id. § 571.208, S4.4.1. Two years later, NHTSA issued a Notice of Proposed Rule Making addressing passenger seating in buses. Bus Passenger Seating and Crash Protection, 38 Fed.Reg. 4776 (Feb. 22, 1973) (to be codified at 49 C.F.R. pt. 571). NHTSA designed the proposed rules to protect bus passengers by requiring higher, stronger, and softer seats that contained the passengers during a crash. Id.

In response to suggestions that seatbelts should be required, NHTSA proposed adding them as an alternate restraint system, replete with a detection system that warned the passenger and the driver -of unfastened seat-belts. Id. The proposed rules were to affect all buses. The following year, however, NHTSA withdrew the proposed standard for mo-torcoaches, determining that seating requirements for intercity and transit buses were not justified from a cost/benefit standpoint, and that seatbelt-usage surveys in intercity buses indicated few passengers would utilize seatbelts if provided.

See School Bus Passenger Crash Protection, 39 Fed.Reg. 27,585 (July 30, 1974). The NHTSA did indicate that it would “propose standards in the future in this area if they are found desirable.” Id MCI draws our attention to several other events of note. In 1992, the chief counsel for NHTSA wrote a letter responding to an inquiry about New York legislation that would require seatbelts in motorcoaches. Letter from Paul Jackson Rice, Chief Counsel, NHTSA to C.N. Littler, Coordinator, Regulatory Affairs, Motor Coach Indus. (Aug. 19, 1992).

He concluded that FMVSS 208 would preempt the New York legislation regarding, as relevant here, intercity buses weighing more than 10,000 pounds. He stated that “NHTSA expressly determined that there is not a safety need for safety belts or another type of occupant crash protection at these [passenger] seating positions.” In so opining, NHTSA’s counsel expressly cited and relied upon the 1974 notice in which NHTSA withdrew its proposed standards for motorcoach seating and seatbelt standards, the latter because of low usage rates.

In 2000, five years after the mo-torcoach here was built, NHTSA’s acting administrator wrote a letter to the chairman of the National Transportation Safety Board (NTSB), which had repeatedly asked NHTSA to study the desirability of seatbelt standards for motorcoaches, noting that motorcoach crashes killed about five people per year. Letter from Rosalyn G. Millman, Acting Adm’r, NHTSA, to Jim Hall, Chairman, NTSB (Mar. 3, 2000). Even so, NHTSA said it would explore ways to study the crashworthiness of mo-torcoaches, including the feasibility and safety of seatbelts, in association with their manufacturers.

Two years later, NHTSA announced a joint public meeting with its Canadian counterpart, Transport Canada, regarding the safety of motorcoaches. See Notice of Public Meeting on Motorcoach Safety Improvements, 67 Fed.Reg. 14,903 (Mar. 28, 2002). NHTSA invited comment on several proposed safety improvements, including limiting the size of glazing materials, introducing roof crush safety standards, requiring side curtain airbags, and requiring seatbelts. Id. at 14,904-05.

Finally, in 2007, NHTSA issued a report recommending passenger seatbelts in mo-torcoaches following research designed to determine the best performance standards for the seatbelt assembly and seat anchorages. See NHTSA, Docket 2007-28793, NHTSA’s Approach to Motorcoach Safety 12-14. Based on this report, the Department of Transportation issued ah action plan for motorcoach safety, in which it proposed to begin rulemaking to require seatbelts in motorcoaches. See U.S. Dep’t of Transp., Motorcoach Safety Action Plan 5 (2009).

In August 2010, NHTSA followed up by publishing a Notice of Proposed Rulemaking (NPRM) that calls for three-point seatbelts for passenger seats in new motorcoaches. See Federal Motor Vehicle Safety Standards; Motorcoach Definition; Occupant Crash Protection, 75 Fed.Reg. 50,958 (Aug. 18, 2010) (to be codified at 49 C.F.R. pt. 571). 2. FMVSS 205■ — Federal Regulation of Glazing Materials FMVSS 205 “specifies requirements for glazing materials for use in motor vehicles.” 49 C.F.R. § 571.205, SI.

It is intended to “reduce injuries resulting from impact to glazing surfaces, to ensure a necessary degree of transparency in motor vehicle windows for driver visibility, and to minimize the possibility of occupants being thrown through the vehicle windows in collisions.” Id. at S2. FMVSS 205 incorporates by reference the standards of the American National Standards Institute (ANSI), specifically the standard for Safety Glazing Materials Z26.1 [hereinafter ANSI/SAE Z26.1-1996]. Id. at S3.2(a).

The parties do not dispute that the standard permitted several kinds of glazing materials, including, as relevant here, laminated or tempered glass, and that MCI complied with the standard in manufacturing the subject bus. See ANSI/SAE Z26.1-1996, T.l (Items 1 & 2) (permitting laminated glass throughout a vehicle and tempered glass anywhere other than the windshield). The ANSI standard notes that “[o]ne safety glazing' material may be superior for protection against one type of hazard, whereas another may be superior against another type.

Since accident conditions are not standardized, no one type of safety glazing material can be shown to possess the maximum degree of safety under all conditions, against all conceivable hazards.” Id. § 2.2. In 1988, NHTSA proposed advanced glazing requirements for passenger vehicles and received numerous comments questioning, among other things, “whether this material would actually increase injuries to belted occupants due to head injury, neck loading, and lacerations.” Withdrawal of Advance Notices of Proposed Rulemaking, 67 Fed.Reg. 41,365, 41,366 (June 18, 2002).

In 1991, Congress mandated that NHTSA initiate rulemaking on rollover protection. Id. As part of this study, NHTSA focused on advanced glazing research as a possible method to reduce passenger ejections. Id. In 2001, Congress directed NHTSA to complete its study of glazing materials, and NHTSA issued a final report on the use of glazing materials to mitigate ejections. Id. at 41,-367. Based on this report, NHTSA decided to terminate rulemaking on the issue of advance glazing, citing safety and cost concerns. Id.

Noting the advent of other ejection mitigation systems (such as side air curtains) and the possibility that advanced side glazing increased the risk of neck injuries in some cases, NHTSA determined that its time and resources were better spent on other projects that focused on developing “more comprehensive, performance-based test procedures.” Id. In the 2007 report on motorcoaches, NHTSA emphasized the importance of roof strength because deformations following a crash compromise the window’s ability to prevent ejections. See NHTSA’s Approach to Motorcoach Safety, supra, at 20.

For that reason, NHTSA concluded that “purs[u]ing the seat belt and roof strength approaches has greater potential for providing improved motorcoach occupant protection than continuing only the glazing/window retention strategy.” Id. In sum, FMVSS 205 reflects NHTSA’s conclusion that no one type of glazing material is superior in all situations, and it leaves to the manufacturers the decision of what kind of material — so long as it is one of the required kinds — to use in a particular setting. B.

The Presumption Against Preemption Before discussing whether preemption applies here, we must first address MCI’s argument that the court of appeals improperly applied a presumption against preemption. MCI contends that Geier rejected any “special burden” beyond the application of ordinary preemption principles. The Plaintiffs counter that the presumption is rooted in federalism, not in GeieP s analysis of the interplay between the Safety Act’s preemption and saving clauses. We agree in principal with the Plaintiffs while recognizing that the exact contours of the presumption are far from clear. See Robert N.

Weiner, The Height of Presumption: Preemption mid the Role of Courts, 32 Hamline L.Rev. 727, 727 (2009) (“Few aspects of Supreme Court jurisprudence are as contradictory and convoluted as the so-called ‘presumption against preemption.’ ”). The United State Supreme Court noted a presumption against preemption in Rice v.

Santa Fe Elevator Corp., grounding it in the states’ police power to regulate for the good of their citizens: “[W]e start with the assumption that the historic police powers of the States were not to be superseded by the Federal Act unless that was the clear and manifest purpose of Congress.” 331 U.S. 218, 230, 67 S.Ct. 1146, 91 L.Ed. 1447 (1947); see also Medtronic, Inc. v. Lohr, 518 U.S. 470, 485, 116 S.Ct. 2240, 135 L.Ed.2d 700 (1996) (“[B]ecause the States are independent sovereigns in our federal system, we have long presumed that Congress does not cavalierly pre-empt state-law causes of action.”).

The presumption is particularly strong when Congress legislates “in [a] field which the States have traditionally occupied.” Rice, 331 U.S. at 230, 67 S.Ct. 1146; see also Bates v. Dow Agrosciences LLC, 544 U.S. 431, 449, 125 S.Ct. 1788, 161 L.Ed.2d 687 (2005) (“In areas of traditional state regulation, we assume that a federal statute has not supplanted state law unless Congress has made such an intention clear and manifest.” (quotation marks omitted)).

Citizens’ health and safety are “ ‘primarily and historically, ... matterfs] of local concern,’ ” and thus states have “ ‘great latitude’ ” to protect “ ‘the lives, limbs, health, comfort, and quiet of all persons.’ ” Lohr, 518 U.S. at 475, 116 S.Ct. 2240 (quoting Hillsborough County v. Automated Med. Labs., Inc., 471 U.S. 707, 719, 105 S.Ct. 2371, 85 L.Ed.2d 714 (1985) and Metro. Life Ins. Co. v. Massachusetts, 471 U.S. 724, 756, 105 S.Ct. 2380, 85 L.Ed.2d 728 (1985)) (alterations in original).

Against this backdrop, the Supreme Court in Geier analyzed whether the Safety Act and FMVSS 208 preempted a common-law tort action in which the plaintiff claimed the automobile manufacturer was liable for failing to install airbags in a 1987 vehicle. 529 U.S. at 865, 120 S.Ct. 1913. The Court first addressed the Safety Act’s express preemption clause and limited its application to state legislative and regulatory enactments, concluding that the Act’s saving clause exempted common-law tort actions from the preemption clause’s scope. Id. at 867-68, 120 S.Ct. 1913.

The Court then rejected the argument that the saving clause foreclosed the operation of ordinary implied preemption principles, including obstacle preemption. Id. at 869, 120 S.Ct. 1913. Further, the majority disagreed with the dissent’s suggestion that the two clauses together created a “special burden” disfavoring preemption. Compare id. at 870-74, 120 S.Ct. 1913, with id. at 898-99, 120 S.Ct. 1913 (Stevens, J., dissenting).

Instead, the majority considered the “language, purpose, and administrative workability” of the statutes and concluded that no reading of the two clauses favored jury-imposed safety standards over federal safety standards with which they actually conflict. Id. at 872-73, 120 S.Ct. 1913. In such a case, the operation of ordinary preemption principles dictates that the state-law standard must give way. We do not read Geie'ds special-burden discussion to undermine the presumption against preemption. The former is rooted in two statutory provisions of the Safety Act, the latter in principles of federalism.

Merely because the Safety Act’s saving clause does not create a special burden disfavoring preemption does not eliminate the respective spheres of state and federal sovereignty, and the presumption simply affords deference to the states’ long-standing rights to protect then* citizens absent a clear directive from Congress. In the end, what the majority said in Geier does not negate the presumption against preemption. What the majority did not say is another matter. MCI correctly notes the Geier majority’s silence regarding the presumption.

The Geier dissent noticed as well and decried the majority’s refusal to apply what the dissent considered to be an “ ‘ordinary experience-proved principle! ] of conflict pre-emption.’ ” Id. at 906-07, 120 S.Ct. 1913 (Stevens, J., dissenting) (quoting id. at 874, 120 S.Ct. 1913); see also Altria Group, Inc. v. Good, 555 U.S. 70, 129 S.Ct. 538, 558, 172 L.Ed.2d 398 (2008) (Thomas, J., dissenting) (noting in a discussion of Riegel v.

Medtronic, Inc., 552 U.S. 312, 128 S.Ct. 999, 169 L.Ed.2d 892 (2008), which involved express preemption, the effect of the majority’s refusal to invoke the presumption: “Given the dissent’s clear call for the use of the presumption against pre-emption, the Court’s decision not to invoke it was necessarily a rejection of any role for the presumption in construing the statute.”). Commentators likewise concluded that the Geier majority upended the normal presumption against preemption. In view of the Supreme Court’s recent statements on the issue, however, we cannot agree. In Wyeth v.

Levine, the Court considered whether federal law preempted, under the actual conflict theory, a Vermont jury’s finding that the manufacturer of the drug Phenergan failed to adequately warn of the risks associated with directly injecting the drug into a patient’s vein. 129 S.Ct. at 1190-91. The Court, relying on Lohr, stated the classic formulation of the presumption against preemption and rejected the dissent’s contention that the presumption should not apply to claims of implied conflict preemption. Id. at 1194-95 & n. 3 (citing Lohr, 518 U.S. at 485, 116 S.Ct. 2240 and Rice, 331 U.S. at 230, 67 S.Ct. 1146).

According to the dissent, the Court had never — prior to Wyeth — definitively applied the presumption in the actual-conflict context. Id. at 1228-29 & n. 14 (Alito, J., dissenting). From the majority’s language in Wyeth, we fail to see how the presumption does not apply to all preemption cases, including implied conflict cases.

Id. at 1194-95 & n. 3 (“For its part, the dissent argues that the presumption against pre-emption should not apply to claims of implied conflict preemption at all, but this Court has long held to the contrary.” (citation omitted)); see also Altria Group, 129 S.Ct. at 543 (“When addressing questions of express or implied pre-emption, we begin our analysis with the assumption that the historic police powers of the States [are] not to be superseded by the Federal Act unless that was the clear and manifest purpose of Congress.” (quotation marks omitted; alteration in original)).

That Geier addressed the Safety Act and Wyeth a different statute is, contrary to MCI’s position, irrelevant. Accordingly, we apply the presumption that Congress did not intend to preempt contrary state law absent evidence that such a result was Congress’s clear and manifest purpose. C. Seatbelts and Federal Preemption Given that no federal safety standard even discusses passenger seatbelts in mo-torcoaches, MCI’s preemption claim is predicated on regulatory silence.

That is, MCI argues that NHTSA’s failure to regulate was deliberate and has the same preemptive force as a regulation forbidding passenger seatbelts. While we agree that regulatory silence can preempt state law in rare occasions, we do not agree that NHTSA’s decision not to require seatbelts in motorcoaches is such an occasion. The Supreme Court has stated that “a federal decision to forgo regulation in a given area may imply an authoritative federal determination that the area is best left wraregulated, and in that event would have as much pre-emptive force as a decision to regulate.” Sprietsma v.

Mercury Marine, 537 U.S. 51, 66, 123 S.Ct. 518, 154 L.Ed.2d 466 (2002) (quoting Ark. Elec. Coop. Corp. v. Ark. Pub. Serv. Comm’n, 461 U.S. 375, 384, 103 S.Ct. 1905, 76 L.Ed.2d 1 (1983) (quotation marks omitted)). In applying this standard, the Court focused on the agency’s explanation for its decision not to regulate. See id. at 66-67, 123 S.Ct. 518. Two cases in which the Court considered the preemptive effect of an agency’s decision not to regulate illustrate the rule’s application.

In Sprietsma, the Supreme Court considered whether a claim that a motor boat should have been equipped with a propeller guard was preempted by the Federal Boat Safety Act of 1971, 46 U.S.C. §§ 4301-4311 (Boating Act). Like the Safety Act, the Boating Act contains both an express preemption clause and a saving clause, which the Court interpreted the same way — the preemption clause only affected state legislation and regulations, and the saving clause preserved common-law tort actions. 537 U.S. at 63, 123 S.Ct. 518.

Finding no express preemption, the Court unanimously rejected the claim that the tort action conflicted with the federal regulatory scheme. Id. at 65, 123 S.Ct. 518. Addressing the boat manufacturer’s argument that the “Coast Guard’s decision not to adopt a regulation requiring propeller guards on motorboats” had preemptive force, the Court responded: “It is quite wrong to view that decision as the functional equivalent of a regulation prohibiting all States and their political subdivisions from adopting such a regulation.” Id.

Rather, that decision “is fully consistent with an intent to preserve state regulatory authority pending the adoption of specific federal standards.” Id. The Coast Guard decided not require propeller guards because doing so was not technically feasible, the cost of retrofitting boats was substantial, and accident data did not support such regulation; as such, the high standard for justifying regulations was not met. Id. at 66, 123 S.Ct. 518.

Given this explanation, the Court found nothing in the Coast Guard’s statement “inconsistent with a tort verdict premised on a jury’s finding that some type of propeller guard should have been installed.” Id. at 67, 123 S.Ct. 518. Similarly, in Freightliner Corp. v. Myrick, the Supreme Court refused to give preemptive force to regulatory silence: “We hold that the absence of a federal standard cannot implicitly extinguish state common law.” 514 U.S. 280, 282, 115 S.Ct. 1483, 131 L.Ed.2d 385 (1995).

The plaintiffs claimed the tractor-trailers that crashed into their vehicles should have been equipped with antilock braking systems (ABS). The Safety Act and NHTSA regulations did not require such braking systems, but the manufacturer argued “that the absence of regulation itself constitutes regulation.” Id. at 286, 115 S.Ct. 1483.

The Court distinguished an earlier case in which it had stated that the failure of federal officials “affirmatively to exercise their full authority takes on the character of a ruling that no such regulation is appropriate or approved pursuant to the policy of the statute.” Id. at 286, 115 S.Ct. 1483 (quoting Ray v. Atl. Richfield Co., 435 U.S. 151, 178, 98 S.Ct. 988, 55 L.Ed.2d 179 (1978) (quotation marks omitted)).

In Ray, Congress had intended to consolidate all regulatory power in the federal government, whereas in Myrick, “NHTSA did not decide that the, minimum, objective safety standard required by [the Safety Act] should be the absence of all standards, both federal and state.” Id. at 286-87, 115 S.Ct. 1483. From these cases, it follows that an agency’s mere decision to leave an area unregulated is not enough to preempt state law. Instead, the agency must, consistent with the authority delegated to it by Congress, affirmatively indicate that no regulation is appropriate.

That is, the agency must state that not only will it leave the area unregulated, it will not allow any regulation in that area as a matter of policy. Unless the agency takes this “further step” to disallow state regulation, its decision not to regulate has no preemptive force. Sprietsma, 537 U.S. at 67, 123 S.Ct. 518. When looking at NHTSA’s comments regarding passenger seatbelts in motor-coaches, we do not find any expression of intent to forbid state regulation. NHTSA proposed a requirement for passenger seatbelts in 1973 along with modified seating standards designed to contain passengers during a crash.

See 38 Fed.Reg. 4776 (Feb. 22, 1973). The following year NHTSA continued to advance such seating standards in school buses, but decided to withdraw the proposed standards for mo-torcoaches and transit buses: “The NHTSA has in fact determined that seating requirements for intercity and transit buses are not justified, based on benefit/cost studies of present seating performance in these buses.... Seat belt usage surveys in intercity buses also indicate that a very low percentage of passengers would utilize seat belts if they were provided....

The NHTSA will, of course, propose standards in the future in this area if they are ■ found desirable.” School Bus Passenger Crash Protection, 39 Fed.Reg. 27,585 (July 30, 1974). This explanation indicates that NHTSA simply determined that the cost of installing seatbelts was not justified given the low usage rates. Nothing in these few sentences evidences an intent to prevent a state jury from concluding that seatbelts in a particular setting were appropriate. The similarity of NHTSA’s decision to the Coast Guard’s decision in Sprietsma not to regulate is compelling.

Like NHTSA, the Coast Guard had considered and rejected a proposed standard intended to increase user safety on the grounds that it was not justified. See 537 U.S. at 66, 123 S.Ct. 518. Both agencies engaged in cost/benefit analyses and concluded that the costs outweighed the benefits based on the relevant data. In Sprietsma, the Supreme Court found no' preemption under these circumstances. On these facts,- we reach the same conclusion.

Because we find no clear and manifest statement in the 1974 withdrawal of the proposed regulations— NHTSA’s final, official statement on the subject before the motorcoach here was manufactured — that NHTSA intended as a matter of policy to forbid any state requirement of passenger seatbelts in motor-coaches, we must conclude that the jury’s verdict in this case is not preempted by federal law. Nor does any later evidence change the analysis.

The letter of NHTSA’s chief counsel in 1992 is unremarkable insofar as it concludes that the Safety Act would preempt nonidentical New York legislation — the Act’s express preemption clause compels such a conclusion. See 49 U.S.C. § 30103(b)(1). But then NHTSA’s counsel opined that the agency had “expressly determined that there is not a safety need for safety belts or another type of occupant crash protection at these seating, positions” and cited the 1974 withdrawal of proposed regulation. Letter from Paul Jackson Rice, Chief Counsel, NHTSA to C.N.

Littler, Coordinator, Regulatory Affairs, Motor Coach Indus. (Aug. 19, 1992). We refuse to give preemptive force to this sentence for two reasons. First, there is nothing in this statement that indicates opposition to seatbelts and an affirmative desire to forbid their use. Second, to hold that NHTSA’s decision not to regulate preempts state law based on a letter written eighteen years after the official decision illustrates the dangers of obstacle preemption grounded in agency musings.

A letter, even by NHTSA’s chief counsel, is not federal law, nor it is particularly persuasive regarding NHTSA’s intent in the 1974 withdrawal. The text of the withdrawal notice is not technical or complex and speaks for itself; a later

interpretation of that text is persuasive only to the extent it is correct. See Skidmore v. Swift & Co., 323 U.S. 134, 140, 65 S.Ct. 161, 89 L.Ed. 124 (1944) (giving weight to an agency’s “rulings,

interpretations and opinions” depending “upon the thoroughness evident in [their] consideration, the validity of [their] reasoning, [their] consistency with earlier and later pronouncements, and all those factors which give [them] power to persuade, if lacking power to control”); cf. Wyeth, 129 S.Ct. at 1201 (giving “some weight” to an agency’s views on the impact of tort law on federal objectives when the subject matter is technical and the relevant history complex). We will not overturn a jury’s verdict on such tenuous grounds.

All of NHTSA’s later statements, from letters to notices of public meetings to reports and action plans, demonstrate the agency’s increased willingness to reconsider its 1974 decision to leave motorcoaehes essentially unregulated. These events culminated in the initiation of rulemaking for the installation of passenger seatbelts on motorcoaehes. The jury anticipated NHTSA’s future actions, and we find no conflict between its verdict and the federal safety standards. MCI argues that the history of NHTSA’s motorcoach regulation points in a very different direction.

It contends that NHTSA chose an alternate method of passenger protection, compartmentalization, see supra note 6, and required seat-belts only for the driver. Further, MCI asserts that three different cases are more relevant than Sprietsma or Myrick: Gei er, Carden v. Gen. Motors Corp., 509 F.3d 227 (5th Cir.2007), and BIC Pen Corp. v. Carter, 251 S.W.3d 500 (Tex.2008). In these latter cases, MCI argues, the agencies carefully weighed competing interests and chose not to require safety devices the plaintiffs wanted.

So too here, MCI continues, NHTSA evaluated the seating designs and desirability of seatbelts in motoreoaehes and purposefully chose a limited scope of regulation. We disagree with MCI’s

interpretation of NHTSA’s actions. In particular, there is nothing in the regulatory history of mo-torcoaches to suggest that NHTSA intended to rely on seating design rather than seatbelts and promulgated standards to that effect. As discussed above, the 1973 notice proposed changing seat design to better protect passengers and also to incorporate seatbelts as an alternative restraint system. In 1974, NHTSA withdrew both proposals, finding that the cost of the methods did not justify their benefits. NHTSA did issue seating standards for school buses, see 49 C.F.R. § 571.222, but not for motoreoaehes.

Thus, it is inaccurate to suggest that NHTSA chose an alternate method to protect passengers; instead, NHTSA chose not to regulate seating design or seatbelts in motorcoach-es, a decision that does not have preemptive force for the reasons explained above. Nor do we find Geier, Carden, or BIC Pen more analogous to the present situation. In Geier, the Supreme Court considered whether FMVSS 208 preempted a claim that a 1987 vehicle should have had airbags. 529 U.S. at 865, 120 S.Ct. 1913.

FMVSS 208, as in effect at the time, “deliberately sought variety — a mix of several different passive restraint systems.” Id. at 878, 120 S.Ct. 1913. NHTSA also chose to gradually phase in the requirements for passive restraints and made them conditional; because seatbelts provided the same or greater benefit at a lower cost, the passive-restraint standard would be rescinded if two-thirds of the states mandated seatbelt use within a certain period of time. Id. at 879-80, 120 S.Ct. 1913.

The reasons for this mix of restraints and gradual phase-in over time reflected NHTSA’s experience and considerations of cost and public acceptance. Even though seatbelts provided the best protection in the event of a crash, most of the public did not wear them at the time. Id. at 877, 120 S.Ct. 1913. While passive restraints such as airbags could provide some of that lost protection, they created problems of their own (e.g., harm to children), cost more than seatbelts, and were viewed with suspicion by the public. Id. at 877-78, 120 S.Ct. 1913.

Thus, rather than simply mandate airbags — a tactic that had failed when NHTSA required seatbelt buzzers and ignition interlocks — NHTSA decided that public “safety would best be promoted if manufacturers installed alternative protection systems in their fleets rather than one particular system in every car.” Id. at 881, 120 S.Ct. 1913 (quotation marks omitted). Because the plaintiffs’ suit required a single standard mandating the installation of airbags, it “presented an obstacle to the variety and mix of devices that the federal regulation sought.” Id.

Because the suit required all of the manufacturer’s cars sold in the District of Columbia to be equipped with airbags, when FMVSS 208 only required ten percent of the manufacturer’s nationwide fleet to be so equipped, it “stood as an obstacle to the gradual passive restraint phase-in that the federal regulation deliberately imposed.” Id. Accordingly, the rule of law advocated by the plaintiffs was preempted by federal law. Id.

In Carden, the Fifth Circuit considered whether FMVSS 208 preempted the plaintiffs’ claim that a car manufacturer should have installed a lap/shoulder seatbelt rather than a lap belt alone in the rear center seat of' a 1999 Pontiac Grand Am. 509 F.3d at 229. Finding Geier directly on point, the court reviewed the history of FMVSS 208 and concluded that NHTSA deliberately gave manufacturers the choice of which kind of seatbelt to install based on specific policy reasons. Id. at 230-31.

These reasons included the technical difficulties associated with installing a shoulder belt in the rear center position, and the greater cost of doing so even when the expected safety benefit was minimal given the low occupancy rate in that seat. Id,, at 231. The Fifth Circuit then distinguished SpHetsma and its analysis of non-regulation by noting the presence of explicit regulation and the choice given therein.- Id. at 232. Finally, in BIC Pen, this Court considered whether federal law preempted a design defect claim regarding a BIC lighter. 251 S.W.3d at 503.

We reviewed the certification process lighters must undergo at the direction of the Consumer Product Safety Commission, the regulatory agency charged with developing safety standards for consumer products. Id. The Commission adopted regulations that require lighters to successfully resist operation by eighty-five percent of a child-test panel. Id.

This number was chosen based on numerous factors, “including child insistence, overall safety, the realities of manufacturing, the variability and randomness of' child testing, the product’s utility, and the importance of customer acceptance.” Id. at 507 (citing 16 C.F.R. § 1210.5(c)). The Commission was particularly concerned that lighters not be too difficult to operate or adults would forgo their use and resort to non-child-resistant lighters or matches, which posed even greater dangers to children. Id.

Because the Commission weighed these factors and struck a delicate balance between them, we concluded that “imposing a common law rule that would impose liability above the federal standard is contrary to the Commission’s plan and conflicts with federal law.” Id. Accordingly, we held that the design defect claim was preempted. Id. at 509. MCI argues these cases illustrate the preemptive effect of an agency’s regulatory action following a careful analysis of various, and often competing, considerations of safety and cost. But MCI misses a telling distinction between these cases and the present one.

In Geier, Carden, and BIC Pen, the agencies actually issued regulations. These regulations balanced competing considerations, as MCI notes, but in each case the courts considered the preemptive effect of a regulation. When an agency issues regulations, a federal law exists with which a state law can conflict. But when the agency chooses not to regulate, there is no preemptive federal law absent a clear and manifest indication of the agency’s intent to forbid all regulation. NHTSA has not taken that further step regarding passenger seatbelts in motor-coaches.

In this respect, the present case is more akin to SpHetsma and Myriclc, in which the Supreme Court found no preemptive intent in the agencies’ non-action. Regulatory silence will not preempt a state law absent a clear and manifest statement of intent to forbid all regulation in that area. Applying this standard to NHTSA’s decision not to require passenger seatbelts in motorcoaches, we find nothing in the regulatory history or agency statements indicative of such intent.

In this lacuna of regulation, the jury’s finding that MCI should have installed seatbelts on its motorcoach does not conflict with any federal law and is not preempted. D. Glazing Materials and Federal Preemption We next consider the effect of FMVSS 205 on the Plaintiffs’ claim that the motorcoach should have had laminated-glass windows. Unlike the seatbelt claim, here NHTSA issued an actual federal safety standard giving manufacturers a choice of glazing materials, including laminated and tempered glass.

The jury determined that MCI should have used laminated glass rather than tempered glass and that this defect caused some of the Plaintiffs’ injuries. The parties agree that FMVSS 205 gives a choice but disagree on its significance. Thus, we must decide if an agency’s deliberate decision to give manufacturers a choice between several different materials, none of which is superior to the others in all circumstances, preempts a jury’s conclusion that another of the required types should have been used. We conclude that it does not.

The Safety Act defines the federal motor vehicle safety standards as “minimum standard[s].” 49 U.S.C. § 30102(a)(9). And state regulation of vehicle safety through common-law tort actions is expressly allowed: “[The Safety Act’s saving clause] preserves those actions that seek to establish greater safety than the minimum safety achieved by a federal regulation intended to provide a floor.” Geier, 529 U.S. at 870, 120 S.Ct. 1913; see also 49 U.S.C. § 30103(e).

Of course, a tort action that seeks to impose a requirement forbidden by a federal standard, or that forbids what the standard requires, is preempted due to the actual conflict. Likewise, a tort action that presents an obstacle to the federal purpose is preempted — Geier is clear that ordinary preemption principles apply. 529 U.S. at 874, 120 S.Ct. 1913. But we must be mindful that Congress generally intended the federal safety standards to set a minimum standard for performance and allowed juries to determine in particular cases if the vehicle manufacturer should have done more.

The text of FMVSS 205 states its threefold purpose — to reduce injuries resulting from impact to glazing surfaces, to provide driver visibility, and to minimize ejections — and incorporates by reference the standards of ANSI/SAE Z26.1-1996. 49 C.F.R. § 571.205, S2, S3.2(a). The ANSI standard delineates the testing requirements for the various glazing materials and allows the use of either laminated glass or tempered glass in vehicle -windows other than the windshield, which must have laminated glass. See ANSI/SAE Z26.1-1996, T.T (Items 1 & 2).

Nothing in the text of FMVSS 205 indicates that it is anything other than a minimum materials standard. In the absence of the standard, manufacturers could use any material allowed by state law; the standard simply limits the range of available choices. MCI argues that the choice of glazing materials is enough to preempt a jury’s finding that a different material should have been used. To be clear, the jury did not find that MCI should have used a glazing material not permitted by FMVSS 205, only that MCI should have used a different glazing material allowed by the standard.

Even such a finding, MCI contends, violates the purpose of FMVSS 205 and the policy-motivated decisions NHTSA made in adopting it. Central to MCI’s argument is its

interpretation of Geier and that opinion’s analysis of FMVSS 208. MCI claims that Geier and lower courts following it have applied FMVSS 208 to preempt tort actions that sought to hold manufacturers liable for not choosing a different safety measure allowed by the standard. Lower courts have, in fact, interpreted Geier in this fashion, but we believe these courts have read Geier too broadly. The Supreme Court in Geier did not condemn a common-law rule merely because it foreclosed a choice under the relevant safety standard.

Rather, the Court emphasized the purpose of the choice: “FMVSS 208 embodies the Secretary’s policy judgment that safety would best be promoted if manufacturers installed alternative protection systems in their fleets rather than one particular system in every car.” 529 U.S. at 881, 120 S.Ct. 1913 (quotation marks omitted). That is, the choice itself furthered the overall goal of safety, allowing manufacturers to experiment with different options and to build the public’s confidence in the new technologies.

The specific reasons for this mix of safety options phased-in over time is discussed above, and there is no doubt that Geier s choice-with-a-purpose reasoning is securely grounded in FMVSS 208 and its particular history. To the extent that other courts have interpreted Geier in the context of FMVSS 208 and NHTSA’s decision to phase-in a mix of passive restraint systems, we agree they are correct.

But when Geier s reasoning is oversimplified to find preemption based on a choice between two safety options and then exported to other safety standards where the unique text and history of FMVSS 208’s passive restraint requirements are not relevant, we must respectfully disagree. Given our

interpretation of Geier, we turn to MCI’s argument that FMVSS 205 represents NHTSA’s deliberate decision to give manufacturers a choice of safety options. That is, MCI seems to contend that NHTSA made a Geier-like policy decision to encourage a range of glazing choices. We do not agree. As noted, FMVSS 205 then and now gives manufacturers a choice of materials and recognizes that no one type is superior in all circumstances. See 49 C.F.R. § 571.205, S3.2(a); ANSI/SAE Z26.1-1996 § 2.2.

In the final rule adopting the new ANSI/SAE standards, NHTSA did not state a positive desire to preserve the use of tempered glass in windows by forbidding contrary state regulation. See 68 Fed.Reg. 43,964. Rather, NHTSA declined to continue rulemaking regarding advanced glazing materials after completing a ten-year study of the subject because other safety measures, such as side air curtains, also helped to mitigate ejections and NHTSA needed to devote its resources to developing standards for them. See 67 Fed.Reg. at 41,366.

NHTSA also cited the costs associated with redesigning vehicles to accept advanced glazing materials and noted that use of these materials may increase the risk of neck injuries. Id. In short, NHTSA extensively studied the issue and did not change the safety standard, which still allows manufacturers to choose among several types of glazing materials. FMVSS 205 is unlike FMVSS 208 and its carefully constructed timetable and mix of safety options. Compare 49 Fed.Reg. 6732, and 68 Fed.Reg. 43,964, with Occupant Crash Protection Final Rule, 49 Fed.Reg. 28,962 (July 17, 1984) (to be codified at 49 C.F.R. pt. 571).

Neither set of standards is, as MCI argues, a “choice among the best available alternatives.” Rather, the former merely narrows the range of manufacturers’ choice of glazing materials from potentially unlimited to a short list. The latter emphasizes the choice among options as an important and integral part of the overall safety scheme. We find nothing in the standard’s text, history, or NHTSA’s comments to indicate that FMVSS 205 is anything other than a minimum standard. As a minimum standard, FMVSS 205 does not preempt the jury’s finding that MCI should have used laminated glass in the motorcoach’s windows.

We are not the first court to examine the preemptive effect of FMVSS 205. In O’Hara v. General Motors Corp., the Fifth Circuit addressed the same issue in the context of a sport utility vehicle. 508 F.3d 753, 755 (5th Cir.2007). The plaintiffs claimed that General Motors should have used advanced glazing materials in the side windows of a 2004 Chevrolet Tahoe instead of tempered glass. Id. The Fifth Circuit considered the text and history of FMVSS 205 as well as NHTSA

interpretations of the standard and general comments on the subject matter of advanced glazing materials. See id. at 759-63. It concluded that FMVSS 205 differs from FMVSS 208 both in text and purpose— specifically that the relevant “factors” of the latter, “detailed implementation time-lines, full vehicle testing procedures and ‘options’ language,” that supported preemption — were “conspicuously absent from FMVSS 205.” Id. at 760. The court found that NHTSA commentary on FMVSS 205 supported “the conclusion that it is a minimum safety standard.” Id. at 761.

Interestingly, the court also considered NHTSA’s decision to terminate rulemaking on advanced glazing materials and analogized that choice to the Coast Guard’s non-action in Sprietsma: We find the parallels between NHTSA’s Withdrawal of Rulemaking and the Coast Guard’s statements in Sprietsma to be compelling. NHTSA’s 2002 Notice of Withdrawal focused on the need to develop experimental standards for new rollover accident technologies. It did not reject advanced glazing as unsafe (indeed, FMVSS 205 continued to require advanced glazing in vehicle windshields).

Like the Coast Guard, the NHTSA Notice of Withdrawal cited cost concerns and minor safety issues to justify the agency’s change in course. And also like the Coast Guard, NHTSA has continued to study advanced glazing as part of its rollover protection program. NHTSA’s Notice of Withdrawal “does not convey an authoritative message of a federal policy against” advanced glazing in side windows. Id. at 762-63. Finding FMVSS 205 to be a minimum safety standard, the court held that the plaintiffs’ negligence and strict liability claims were not preempted. Id. at 763.

The Supreme Court of West Virginia also considered whether FMVSS 205 preempted a claim that a vehicle manufacturer should have used advanced glazing materials rather than tempered glass in the side window of a sport utility vehicle. Morgan v. Ford Motor Co., 224 W.Va. 62, 680 S.E.2d 77, 81 (2009). The court examined Geier, O’Hara, and Wyeth at some length, see id. at 88-93, and specifically rejected the

interpretation of Geier that we adopt, instead holding that a regulation giving a choice between different materials preempts a tort action premised on the superiority of one option, id. at 94-95. The court reasoned that “because the NHTSA made a public policy decision to not mandate advanced glazing in side windows because of safety concerns that advanced glazing has a slightly increased risk of neck injuries,” it was compelled to find Geier directly applicable. Id. at 94. As our analysis makes plain, we agree with O’Hara.

The Morgan court errs by concluding that merely because NHTSA chose not to require something for policy reasons, states may not do so as well. As discussed above, the rule in Sprietsma regarding the preemptive force of regulatory non-action is not so broad, particularly when NHTSA still permits manufacturers to choose what it would not require. There is no evidence that NHTSA intended to disallow states from requiring the use of advanced glazing materials in side windows. Nor do we agree with the Morgan court’s broad

interpretation of Geier, that preemption is mandated whenever an agency makes a considered decision to preserve the status quo and its range of choices. As the Solicitor General has said: “Manufacturers always have the ‘option’ of exceeding a minimum safety standard when NHTSA has decided not to mandate a more stringent alternative because of considerations of cost or feasibility — as NHTSA did in this case and, indeed, often does in considering regulatory alternatives.

But if such an ‘option’ alone were enough to trigger federal preemption under Geier, the Safety Act’s savings clause would be greatly undermined.” Brief for the United States • as Amicus Curiae at 15, Williamson v. Mazda Motor of Am., Inc., No. 08-1314. We agree. Attributing preemptive intent to every deliberate agency decision runs afoul of Congress’s choice to define the safety standards as minimum standards and its clear decision to allow juries a place in developing common-law rules that exceed the federally defined floor.

We hold that FMVSS 205 is a minimum standard that merely limits the possible glazing materials a manufacturer may choose for incorporation in its vehicles. As such, the jury’s finding that MCI should have used a different glazing material in the motorcoach here presents no obstacle to the accomplishment and execution of a federal policy. Thus, we hold that the jury’s verdict in favor of the Plaintiffs is not preemptéd by federal law. III. Settling Persons We finally consider whether Central Texas and the bus driver are settling persons under

Chapter 33 of the Texas Civil Practice and Remedies Code. The 1995 version of

Chapter 33 provided a proportionate responsibility framework for apportioning percentages of responsibility in the calculation of damages. At that time,

section 33.003 required a trier of fact to determine the percentage of responsibility for each claimant, each defendant, each settling person, and each responsible third party who had been properly joined. Tex. Civ. Prac. & Rem.Code § 33.003.

Section 33.013 provided, with some exceptions, that a defendant was only liable for the percentage of responsibility found by the trier of fact, unless the percentage exceeded fifty percent. Id. § 33.013(a); F.F.P. Operating Partners, L.P. v. Duenez, 237 S.W.3d 680, 687 (Tex.2007). If a defendant’s percentage of liability exceeded fifty percent, the defendant was jointly and severally liable for the entirety of the damages recoverable by the claimant. Tex. Civ. Prac. & Rem.Code § 33.013(b); Duenez, 237 S.W.3d at 687. Relevant to our determination of this issue is the definition of a “settling person.” While “settle” and “settlement” were not defined in

Chapter 33, see C & H Nationwide, Inc. v. Thompson, 903 S.W.2d 315, 319 (Tex.1994), “settling person” was. A “settling person” was “a person who at the time of submission has paid or promised to pay money or anything of monetary value to a claimant at any time in consideration of potential liability ... with respect to the personal injury, property damage, death, or other harm for which recovery of damages is sought.” Tex. Civ. Prac. & Rem.Code § 33.011(5). Here, the trial court refused MCI’s request to include Central Texas as a “settling person” in a proportionate responsibility question in the jury charge.

Whether the trial court erred in this ruling is crucial: a finding by the jury that Central Texas had a percentage of responsibility as a settling person could potentially have reduced MCI’s percentage of liability to fifty percent or less, thus limiting the amount of damages MCI would have had to pay. Our analysis of whether Central Texas is a settling person under

Chapter 33 turns on a question of statutory construction. A question of statutory construction is a legal one which we review de novo, “ascertaining and giving effect to the Legislature’s intent as expressed by the plain and common meaning of the statute’s words.” Duenez, 237 S.W.3d at 683 (citing Tex. Dep’t of Transp. v. City of Sunset Valley, 146 S.W.3d 637, 642 (Tex.2004)); see also Leland v. Brandal, 257 S.W.3d 204, 206 (Tex.2008) (citing Nat’l Liab. & Fire Ins. Co. v.

Allen, 15 S.W.3d 525, 527 (Tex.2000)) (observing the necessity of a court determining and giving effect to the Legislature’s intent in construing a statute). We first look at “the statute’s language to determine that intent, as we consider it ‘a fair assumption that the Legislature tries to say what it means, and therefore the words it chooses should be the surest guide to legislative intent.’ ” Brandal, 257 S.W.3d at 207 (quoting Fitzgerald v. Advanced Spine Fixation Sys., Inc., 996 S.W.2d 864, 866 (Tex.1999)).

Thus, we consider the statute’s plain and common meaning, and do not “look to extraneous matters for an intent the statute does not state.” Allen, 15 S.W.3d at 527. As noted, the applicable version of

Chapter 33 defined “settling person” as a person who, at the time of submission, has paid or promised to pay anything of value to a claimant in consideration of liability. C & H Nationwide, 903 S.W.2d at 320. We apply the plain meaning of these words to the facts before us. Here, Central Texas had filed for bankruptcy, but its insurer deposited, on the bankruptcy court’s order, the limits of its policy into the court’s registry for payment to the Plaintiffs. Central Texas also agreed to pay the claimants $7,000 a year for five years.

There is no question that the claimants who chose to obtain funds from the Apportionment Plan “settled” with Central Texas — they received money in exchange for a release of Central Texas’s liability. But the Plaintiffs chose to participate in the Litigation Plan, which they claim could not constitute a settlement due to its uncertain and adversarial nature at the time of submission in the trial court.

As the court of appeals observed, all the parties in the bankruptcy court engaged in extensive negotiations over the division of the funds among the litigants, and Central Texas and its insurer did not oppose tendering the funds into the court’s registry or participating in mediation. Thus, Central Texas definitely paid something of monetary value to the claimants, including the Plaintiffs — $5 million in insurance proceeds, plus $7,000 annually — in respect to its potential liability, and the Plaintiffs negotiated for, and were the ultimate recipients of, some of these proceeds.

Central Texas was further discharged from tort liability in the bankruptcy court on approval of its reorganization plan, which the Plaintiffs approved. Although the Litigation Plan was arguably not styled a settlement by the claimants or bankruptcy court, it had all semblances of a settlement. See Ratcliff v. Fibreboard Corp., 819 F.Supp. 584, 588-89 (W.D.Tex.1992) (concluding that agreement was a final settlement under

Chapter 33, despite language in the agreement to the contrary). We hold, under these facts, that Central Texas is a settling person, and the trial court should have submitted a question to the jury concerning Central Texas’s proportionate responsibility as such. The Plaintiffs advance several arguments supporting their position that Central Texas cannot be considered a settling person under the applicable version of

Chapter 33. We address each in turn. First, the Plaintiffs contend that the uncertain and adversarial nature of the Litigation Plan renders the Plan a form of litigation. One group of claimants — those who elected to participate in the Apportionment Plan — did immediately collect their proportion of the insurance funds, while those who elected to participate in the Litigation Plan deferred their receipt of the funds. Our inquiry, then, is whether electing to defer the collection of funds permits the Plaintiffs to deny Central Texas’s status as a settling person. We con-elude it does not.

The definition of settling person requires only a promise to pay, not an actual payment, in consideration for a release from liability. See C & H Nationwide, 903 S.W.2d at 320. Here, there was, in essence, a promise to pay in consideration for Central Texas’s release from liability. While the Plaintiffs chose to present arguments in a hearing before a special judge, the Litigation Plan also allowed the Plaintiffs to “agree at any time to approve full or partial distribution of the Litigation Funds to any or all participants,” even without going before the special judge.

There was never a real possibility that the Plaintiffs would not recover some amount of proceeds from the Litigation Fund: the Plaintiffs not only had the option of receiving the proceeds in the Litigation Fund at any time, but the Litigation Plan contained no provision for the return of any nonawarded proceeds to Central Texas or its insurer. Further, even though the Plaintiffs chose not to approve distribution from the Litigation Fund before the hearing in front of the special judge, the hearing was not adversarial in nature and the end result of the hearing was certain before it began.

Central Texas was not present at the hearing and, of course, had no reason to be present since it had irrevocably paid money into the court’s registry in exchange for a release from liability. Thus, the hearing lacked the adversarial nature the Plaintiffs contend existed. The Plaintiffs argue that the amount each claimant might have received following the hearing was uncertain.

But the Litigation Plan capped the amount of each claimant’s recovery at 110% of that determined by the mediator, and each participant in the Litigation Plan received within two percent of the amount determined by the mediator, with one exception. Even if the exact amount each claimant might have received was uncertain, there was little uncertainty that each Plaintiff was going to receive some amount. The statute does not require certainty in the actual amount of the money or thing of monetary value — just the promise that the person will receive something of value. Tex. Civ.

Prac. & Rem.Code § 33.011(5); see also Gilcrease v. Garlock, Inc., 211 S.W.3d 448, 455 (Tex.App.-El Paso 2006, no pet.) (holding settlement agreement was not contingent because the defendants made an unconditional promise to pay). Second, the Plaintiffs contend that Central Texas’s release from liability in the bankruptcy court was involuntary under federal bankruptcy law, and thus cannot constitute a settlement. See In re Arrowmill Dev. Corp., 211 B.R. 497, 503 (Bankr.D.N.J.1997).

But to hold that the definition of “settling person” requires a voluntary release of claims would compel us to insert language into the statute that is not there. See Fortis Benefits v. Cantu, 234 S.W.3d 642, 649 n. 41 (Tex.2007) (noting that courts “should not by judicial fiat insert non-existent language into statutes”). In any event, it is inaccurate to describe Central Texas’s release from liability as involuntary.

Central Texas and its insurer engaged in negotiations concerning not only the money to be placed in the bankruptcy court’s registry and the manner in which funds were to be allocated, but also whether or not Central Texas was to be released from liability. When the Plaintiffs voted in favor of approving Central Texas’s reorganization plan, they consented to release Central Texas from liability. Third, the Plaintiffs contend that there was uncertainty “at the time of submission,Tex. Civ.

Prac. & Rem.Code § 33.011(5) (requiring the payment or promise to pay something of monetary value to have occurred “at the time of submission” in order for a person to be considered a “settling person”) (emphasis added). But the fact that the Plaintiffs had not actually presented their cases to the special judge at the time of submission does not change whether or not the settlement was certain.

At the time of submission in the trial court, Central Texas and its insurer had already deposited funds in the bankruptcy court’s registry, the Litigation Plan had been in place for over two years, and the Plaintiffs had the option of requesting disbursement of their proportionate shares of the Fund. Fourth, the Plaintiffs argue that the proper vehicle for proportioning some share of Central Texas’s potential liability would have been through its joinder as a responsible third party. Under the applicable version of

Chapter 33, a responsible third party was defined as a person (1) over whom the court can exercise jurisdiction, (2) who was not sued by the claimant, and (3) who is or may be liable for all or part of the damages claimed against the named defendant. Tex. Civ. Prac. & Rem. Code § 33.011(6).

The term specifically excluded “a person or entity that is a debtor in bankruptcy proceedings or a person or entity against whom this claimant’s claim has been discharged in bankruptcy, except to the extent that liability insurance or other source of third party funding may be available to pay claims asserted against the debtor.” Id.

The Plaintiffs contend that since a debtor in bankruptcy without the availability of liability insurance was excluded from the definition of responsible third party, it follows that a debtor in bankruptcy with the availability of liability insurance — like Central Texas — is a responsible third party. Assuming that Central Texas could properly be characterized as a responsible third party, nothing in

Chapter 33 suggested that a party’s identification as a responsible third party or settling person is an either-or proposition. A person could potentially fall within the

definitions of both settling person and responsible third party, and the Legislature included no language in

Chapter 33 indicating otherwise. See Brandal, 257 S.W.3d at 206 (noting that the Legislature “tries to say what it means, and therefore the words it chooses should be the surest guide to legislative intent” (citing Fitzger ald, 996 S.W.2d at 866)); Lee v. City of Houston, 807 S.W.2d 290, 295 (Tex.1991) (observing that a court may not “judicially amend a statute and add words that are not implicitly contained in the language of the statute”). The Legislature did not limit the designations of settling person and responsible third party as exclusive, and we will not read such a limitation into the statute.

See C & H Nationwide, 903 S.W.2d at 320 (concluding that the definition of “settling person” is not limited to those who fully resolve all claims against them and includes those who settle only partially); City of Rockwall v. Hughes, 246 S.W.3d 621, 629 (Tex.2008) (noting that a statute is to be interpreted as written, unless the context requires otherwise or the construction would lead to an absurd result).

Fifth, the Plaintiffs contend that to find Central Texas a settling person will confuse settlement jurisprudence and over-broadly designate a person as settling any time the person deposits funds in a court’s registry. We disagree with this assertion. We do not hold that a deposit of funds in a court’s registry is always akin to a settlement. Rather, it is so when, as here, the negotiations and terms of an agreement in every way resemble a settlement. When this is the.case, we must define it as what it is — a settlement.

As discussed before, Central Texas and its insurer did not simply deposit funds in the bankruptcy court’s registry. They negotiated the terms of the deposits with the claimants and entered into a mediated agreement where the claimants could either accept the funds immediately or defer acceptance to a later date, and Central Texas was released from liability. Our holding will also have no adverse effect on parties in bankruptcy proceedings or in settlement negotiations.

To the contrary, parties in bankruptcy and related litigation will know that if they enter into an agreement that in all respects resembles a settlement, they will be deemed settling persons in related state litigation for purposes of

Chapter 33. Finally, the Plaintiffs point to statements made both in documents in the bankruptcy court and by the bankruptcy court judge suggesting that the Litigation Plan was not a settlement under

Chapter 33. In his dissent, Chief Justice Jefferson also relies on these statements as indication that the Litigation Plan was not a settlement. But these statements were not part of the bankruptcy court’s holdings. As the court of appeals observed, these were merely ipse dixit statements that, in any event, have no bearing on our

interpretation of Texas law. See Allen, 15 S.W.3d at 527 (observing that we consider the statute’s plain and common meaning and do not “look to extraneous matters for an intent the statute does not state”); cf. In re W.E.R., 669 S.W.2d 716, 716 (Tex.1984) (per curiam) (holding that appellate court may not accept a trial judge’s oral comments as findings of fact or conclusions of law). Chief Justice Jefferson asserts that we improperly construe the circumstances under which the Plaintiffs pursued their claims in the bankruptcy court. We respectfully disagree.

As discussed above, the portion of the Apportionment Plan concerning the Litigation Plan specifically provided that the Litigation Plan participants could agree at any time to distribution of Litigation Funds to any or all participants, provided that it was done in writing by all participants in the Litigation Plan. Although the Plaintiffs did not ultimately exercise this option, it was available to them at the time of submission in the trial court.

Thus, at the time of submission, not only had Central Texas unilaterally deposited funds in the bankruptcy court registry for future distribution to the claimants, but the Plaintiffs — and other Litigation Plan participants — had the option of receiving those funds at any time. Moreover, even though the Plaintiffs ultimately decided to pursue the “litigation” option, the record belies a description of the hearing before the special judge as adversarial or contingent.

The claimants crafted the Litigation Plan after extensive negotiations and, as previously mentioned, could not individually recover more than 110% of the mediator’s allocation. While it is true that the Litigation Plan did not set a floor on each claimant’s possible individual recovery, there was never a real possibility that any of the Plaintiffs — all victims of the bus accident and their families — would receive none of the bankruptcy court funds, given that the hearing was uncontested and the Litigation Plan contained no provision for the return of undis-bursed funds to Central Texas.

This is not a retroactive view of the proceedings: because Central Texas’s reorganization plan had previously been approved, Central Texas had no reason to appear at the hearing. Of course, the dissent correctly notes that the actual results of the hearing before the special judge are not relevant since we must determine settlement “at the time of submission.” But the results — with all but one participant receiving within two percent of the amount assigned by the mediator — do highlight the fact that this was not actual, adversarial litigation with an uncertain outcome.

There is further no indication in the record that other “potential parties” were entitled to the Litigation Funds in lieu of the Litigation Plan participants. Rather, the bankruptcy judge stated that other parties might be entitled to those funds if the Litigation Plan participants were not for whatever reason.

At the time of submission, Central Texas had deposited the limits of its insurance policy in the bankruptcy registry and had been released from liability, and the Litigation Plan participants had the option of either receiving their proportionate shares of the funds or bringing their cases before a special judge in a nonadversarial hearing. Despite how the bankruptcy court or the Plaintiffs might describe these events, we conclude they constitute a settlement under

Chapter 33.

Chapter 33 expresses the Legislature’s intent to hold defendants responsible for only their own conduct. Its purpose is to hold each person “responsible for [the person’s] own conduct causing injury.” Duenez, 237 S.W.3d at 690. “This is consistent with a fundamental tenet of tort law that an entity’s liability arises from its own injury-causing conduct.” Id. We therefore hold that the trial court erred in refusing to submit a question to the jury concerning Central Texas’s proportionate responsibility as a settling person. IV.

Conclusion For the foregoing reasons, we affirm the court of appeals’ judgment and remand the case to the trial court for further proceedings consistent with this opinion. Chief Justice JEFFERSON filed an opinion, dissenting in part. Justice GREEN did not participate in the decision. . Motorcoach, intercity bus, and bus are used interchangeably in this opinion, but are distinguished from transit buses. See Notice of Public Meeting on Motorcoach Safety Improvements, 67 Fed.Reg. 14,903 (Mar. 28, 2002). .

To be clear, the bus crash victims who submitted claims in the bankruptcy proceeding are a larger group than the Plaintiffs here. Some of these claimants opted for compensation under the Apportionment Plan, leaving about half of the insurance proceeds for distribution to those who chose to participate in the Litigation Plan. . One participant received less than the amount assigned by the mediator in line with the jury findings in the trial court. .

Justice Thomas has gone so far as to question whether implied obstacle preemption can be reconciled with the Supremacy Clause’s mandate that only federal law can preempt state law: "Congressional and agency musings, however, do not satisfy the Art. I, § 7 requirements for enactment of federal law and, therefore, do not pre-empt state law under the Supremacy Clause.” Wyeth, 129 S.Ct. at 1207 (Thomas, J., concurring in the judgment).

Instead, Justice Thomas contends, "[p]re-emption must turn on whether state law conflicts with the text of the relevant federal statute or with the federal regulations authorized by that text.” Id. at 1208. However, the United States Supreme Court continues to uphold and apply obstacle preemption based on federal purpose, and we are bound to follow. . For buses manufactured between January 1, 1972 and September 1, 1990, manufacturers actually had a choice even for the driver position. § 571.208, S4.4.1. They could design a "complete passenger protection s

[…]