Summary
holding that the Court of Appeals misapplied the expressio unius canon in construing provisions of the Insurance Code that permit no-fault insurers to seek reimbursement for payment of some benefits as implicitly excluding any other reimbursement mechanism
Summary of this case from Esurance Property & Casualty Insurance Co. v. Michigan Assigned Claims PlanOpinion
No. 160242
05-27-2021
Keith BRONNER, Plaintiff, and Angels With Wings Transport, LLC, Intervening Plaintiff, v. CITY OF DETROIT, Defendant/Third-Party Plaintiff-Appellant, and GFL Environmental USA Inc., f/k/a Rizzo Environmental Services, Inc., Third-Party Defendant-Appellee.
Charles N. Raimi for the City of Detroit. Cardelli Lanfear, PC, Royal Oak (by Anthony F. Caffrey III and Thomas G. Cardelli ) for GFL Environmental USA Inc.
Charles N. Raimi for the City of Detroit.
Cardelli Lanfear, PC, Royal Oak (by Anthony F. Caffrey III and Thomas G. Cardelli ) for GFL Environmental USA Inc.
BEFORE THE ENTIRE BENCH
Clement, J. In this case, we consider whether a no-fault insurer—or, as here, a self-insurer—may legally contract with a vendor for indemnification of the no-fault insurer for the cost of no-fault benefits that the insurer is obliged by law to pay when the vendor's negligence caused the injury for which the benefits are compensation. We conclude that such an agreement is legal and reverse the contrary conclusion of the Court of Appeals.
I. FACTS AND PROCEDURAL HISTORY
On September 25, 2014, Keith Bronner was a passenger on a bus operated by the City of Detroit. The bus was in an accident with a garbage truck operated by GFL Environmental USA Inc. The city self-insures its fleet of buses under MCL 500.3101(5), and Bronner consequently made a claim with the city for personal protection insurance (PIP) benefits under MCL 500.3107. The city initially paid about $58,000 in benefits to Bronner; but eventually the relationship broke down, and Bronner sued the city in September 2015.
At the time of the accident, GFL was known as Rizzo Environmental Services, Inc.
At the time of these events, the relevant provision was found at MCL 500.3101(4) ; it was renumbered with the enactment of 2019 PA 21. Because 2019 PA 21 does not affect this dispute, references in this opinion are to the current version of the statute.
GFL's garbage truck was operating under a contract that GFL had signed with the city in February 2014. Section 9.01(a) of that agreement provided that GFL
agree[d] to indemnify, defend, and hold the City harmless against and from any and all liabilities, obligations, damages, penalties, claims, costs, charges, losses and expenses ... that may be imposed upon, incurred by, or
asserted against the City ... to the extent caused by ... [a]ny negligent or tortious act, error, or omission attributable in whole or in part to [GFL] or any of its Associates[.]
Shortly after Bronner sued the city, the city filed a third-party complaint against GFL, invoking this indemnification agreement. In June 2016, GFL moved for summary disposition, arguing that the city was "attempting to circumvent the explicit requirements of the No-Fault Act[ ] by improperly shifting its burden onto [GFL] through language found in an unrelated service contract between Detroit and [GFL], which clearly violates public policy and the legislative intent behind the No-Fault Statute." The trial court denied this motion and instead granted summary disposition in favor of the city. In February 2017, the city reached a settlement with Bronner, and the trial court then ordered GFL to pay the city $107,529.29 to cover the PIP benefits paid by the city, plus certain other expenses.
MCL 500.3101 et seq.
This sum included both the city's settlement with Bronner and its settlement with Angels with Wings Transport, LLC, which had provided transportation services to Bronner and intervened as a plaintiff to make its own recovery.
In the Court of Appeals, GFL renewed its argument that the indemnification agreement circumvented the Insurance Code's no-fault rules and was therefore void. The Court of Appeals agreed and reversed in an unpublished opinion. The Court of Appeals panel emphasized the comprehensive nature of the no-fault system, which includes only a few explicit mechanisms by which a no-fault insurer may recover the cost of benefits paid out. The Court accepted the negative implication that, by stating these options in the no-fault act, the Legislature had denied the availability of any other options. The panel therefore concluded that the indemnification agreement was unenforceable. The city then filed an application for leave to appeal in our Court, and we ordered argument on that application. Bronner v. Detroit , 505 Mich. 1139, 944 N.W.2d 722 (2020).
MCL 500.100 et seq.
Bronner v. Detroit , unpublished per curiam opinion of the Court of Appeals, issued July 9, 2019 (Docket No. 340930), 2019 WL 3000915.
II. STANDARD OF REVIEW
The question before the Court is not the meaning of the indemnification agreement between the city and GFL as such. GFL's argument in this Court does not concern the proper interpretation of the parties’ contract, and GFL does not argue that the indemnification sought by the city is beyond the scope of that contract. Rather, the question is whether the Insurance Code precludes the contract provision at issue. In other words, the question is whether the provision "runs afoul of the public policy of the state" in the form of "the policies that ... are reflected in ... our statutes," Terrien v. Zwit , 467 Mich. 56, 66-67, 648 N.W.2d 602 (2002), such as the Insurance Code. Whether a contract provision is invalid on these grounds is a question of law subject to de novo review. Id. at 61, 648 N.W.2d 602. "This Court [also] reviews the grant or denial of summary disposition de novo to determine if the moving party is entitled to judgment as a matter of law." Maiden v. Rozwood , 461 Mich. 109, 118, 597 N.W.2d 817 (1999).
III. ANALYSIS
We have held that " ‘[t]he general rule [of contracts] is that competent persons shall have the utmost liberty of contracting and that their agreements voluntarily and fairly made shall be held valid and enforced in the courts.’ " Terrien , 467 Mich. at 71, 648 N.W.2d 602, quoting Twin City Pipe Line Co. v. Harding Glass Co. , 283 U.S. 353, 356, 51 S. Ct. 476, 75 L. Ed. 1112 (1931). Of course, where there are " ‘definite indications in the law’ " of some contrary public policy, Terrien , 467 Mich. at 68, 648 N.W.2d 602, quoting Muschany v. United States , 324 U.S. 49, 66, 65 S. Ct. 442, 89 L. Ed. 744 (1945), the contract provision must yield to public policy. As the Court of Appeals noted here, however, there is no provision of the Insurance Code which expressly prohibits the sort of indemnification agreement at issue. Even so, the Court of Appeals drew inferences from the comprehensive nature of the no-fault system that we must assess.
No-fault insurance in Michigan is "a comprehensive scheme of compensation designed to provide sure and speedy recovery of certain economic losses resulting from motor vehicle accidents." Belcher v. Aetna Cas. & Surety Co. , 409 Mich. 231, 240, 293 N.W.2d 594 (1980). "In general, where comprehensive legislation prescribes in detail a course of conduct to pursue and the parties and things affected, and designates specific limitations and exceptions, the Legislature will be found to have intended that the statute supersede and replace the common law dealing with the subject matter." Millross v. Plum Hollow Golf Club , 429 Mich. 178, 183, 413 N.W.2d 17 (1987). Although Millross was a case about the dramshop act, we have applied this same principle in the no-fault context. In particular, the Court of Appeals drew upon our line of cases construing the comprehensive nature of the no-fault law as prohibiting certain shifts of liability for no-fault benefits to invalidate the indemnification agreement at issue. The first case in this line is Citizens Ins. Co. of America v. Federated Mut. Ins. Co. , 448 Mich. 225, 531 N.W.2d 138 (1995). In that case, a car dealership gave a customer a "loaner" automobile while the dealership was working on the customer's own vehicle. The customer was later in a serious accident. On appeal in this Court, the legal question was which insurer was responsible under MCL 500.3131 to pay residual personal injury benefits: the insurer of the car dealership (as the owner of the vehicle) or the customer (as the operator of the vehicle). The insurance policy issued to the car dealership by its insurer stated that the insurer would not consider as an "insured" anyone to whom the dealership had loaned the vehicle unless that individual was uninsured or underinsured. We held that the dealership's insurer could not, in its policy, unilaterally shift liability for no-fault benefits to fully insured borrowers of the dealership's vehicle because it violated MCL 500.3101(1), which requires the owner of a vehicle to maintain security for residual liability insurance. The policy exclusion was therefore void, and the dealership's insurer had to pay benefits.
The facts of Citizens appear in neither this Court's opinion nor the majority opinion in the Court of Appeals, but rather the dissenting opinion in the Court of Appeals. See Citizens Ins. Co. v. Federated Mut. Ins. Co. , 199 Mich. App. 345, 348, 500 N.W.2d 773 (1993) ( Neff , J., dissenting).
We extended Citizens Ins. Co. in State Farm Mut. Auto. Ins. Co. v. Enterprise Leasing Co. , 452 Mich. 25, 549 N.W.2d 345 (1996). There, we held that when an automobile is rented out, the lessor of the vehicle may not enforce a lease condition shifting responsibility to the lessee's no-fault insurer to provide mandatory no-fault benefits if an accident occurs—even if the lessee agreed to this lease condition. Id. at 27-28, 35, 549 N.W.2d 345. We offered various reasons for this conclusion, but one, which the Court of Appeals referred to here, was that the intent of the no-fault system is to hold the owner rather than the operator of a vehicle primarily responsible for paying no-fault benefits, and it would subvert that intent to switch that responsibility:
The driver cannot defeat the provisions of the no-fault act by stating that the owner need not pay insurance. Because the driver cannot bind the driver's insurer, a driver who agreed to shift coverage would remain solely liable for damages caused by use of the vehicle. The rental car would be left uninsured under the terms of the rental agreement stating that the owner provides no insurance. This lack of coverage violates the no-fault act. Even though an injured party could attempt to obtain compensation from the driver, the no-fault act is intended to protect injured parties from having to pursue such suits. Even if the driver qualified as self-insured, we would not allow the rental car companies to avoid the Legislature's intent that a vehicle owner be primarily responsible for providing coverage. Just as the car rental company cannot shift liability to a driver's insurer, it cannot shift liability to a driver personally. Either shift of responsibility away from the owner would violate the act because it requires owners to provide primary coverage. [ Id. at 35-36, 549 N.W.2d 345.]
On the other hand, we gestured toward a limit to the principle established in Citizens Ins. Co. and State Farm in Universal Underwriters Ins. Co. v. Kneeland , 464 Mich. 491, 628 N.W.2d 491 (2001). In Kneeland , as in Citizens Ins. Co. , a car dealership loaned an automobile to a customer while it was working on her vehicle. Id. at 493, 628 N.W.2d 491. The customer signed an agreement when she borrowed the vehicle in which she "agree[d] to assume all responsibility for damages while [the] vehicle [was] in h[er] possession." Id. (emphasis omitted). While she was driving the vehicle, she was in an accident causing more than $3,700 in damage to the vehicle. Id. The dealership and its insurer paid for appropriate repairs but then sued the customer to enforce the agreement she had made when she borrowed the vehicle, seeking compensation for the cost of the repairs. Id. We expressed concern that the term "damages" in the agreement "could refer to any harm caused to a third party's person or property, i.e., it could reach damages for which no-fault insurance coverage is mandatory." Id. at 496, 628 N.W.2d 491. Citing State Farm , we acknowledged that "[a] shift of liability to that extent might contravene the no-fault act." Id. at 496-497, 628 N.W.2d 491. That said, what was sought under the contract in Kneeland was nonmandatory collision coverage, which took Kneeland outside the rule from Citizens Ins. Co. and State Farm , and we therefore concluded that "the contract thus does not shift liability for damages that may not legally be reallocated." Id. at 498, 628 N.W.2d 491.
The Court of Appeals panel here construed the indemnification provision at issue as a variation on the sort of liability-shifting that these cases have prohibited. In particular, it emphasized certain hedging language from our Kneeland opinion. In interpreting the word "damages," which the vehicle borrower agreed to accept responsibility for in the Kneeland contract, we observed that it could encompass mandatory no-fault benefits and, citing State Farm , we noted that such a shift of liability might violate the Insurance Code. Id. at 496-498. We stated:
See Bronner , unpub. op. at 5-6.
We express no view regarding whether State Farm would control the legality of the contract [in Kneeland ]. Th[e] agreement and the one addressed in State Farm are arguably different in scope and effect. We merely observe that an argument is available that the parties’ agreement, if it reaches beyond optional collision damages, is illegal. [ Id. at 497 n. 3, 628 N.W.2d 491.]
The Court of Appeals panel concluded that this left open whether benefits that the no-fault law requires to be paid out could be shifted and that the existence of various reimbursement mechanisms for no-fault insurers under the statute implicitly precluded the enforceability of an indemnification agreement such as the one at issue.
Bronner , unpub. op. at 6.
The Court of Appeals’ analysis of Kneeland is flawed, however, as it does not read Kneeland in the context of Citizens Ins. Co. and State Farm , which came before Kneeland . This is best demonstrated by reviewing Cruz v. State Farm Mut. Auto. Ins. Co. , 466 Mich. 588, 648 N.W.2d 591 (2002). The insurance policy in Cruz made payment of no-fault benefits contingent on the injured person submitting to an "examination under oath" (EUO), id. at 590, 648 N.W.2d 591, and the question was whether this provision was compatible with the Insurance Code's requirement that no-fault insurers pay benefits "within 30 days after an insurer receives reasonable proof of the fact and of the amount of loss sustained," MCL 500.3142(2). See Cruz , 466 Mich. at 593-594, 596, 648 N.W.2d 591. Because "the no-fault act contains no reference either allowing or prohibiting examinations under oath," we had to "determine whether, given this silence, the inclusion of examination under oath provisions in no-fault automobile insurance policies is allowed." Id. at 594, 648 N.W.2d 591. We held that the parties could not "contract out of the statutory duty imposed on the insurer to pay benefits within thirty days of receipt of the fact and of the amount of the loss sustained by agreeing that no benefits are due until an EUO is given by the insured[.]" Id. at 595, 648 N.W.2d 591. Drawing on Kneeland , we sought to harmonize the EUO requirement in Cruz with the Insurance Code and declined to hold that EUOs intrinsically violate it. Instead, we held that EUOs are permissible "when used to facilitate the goals of the [no-fault] act and when they are harmonious with the Legislature's no-fault insurance regime," such as if they are "designed only to ensure that the insurer is provided with information relating to proof of the fact and of the amount of the loss sustained—i.e., the statutorily required information on the part of the insured." Id. at 598, 648 N.W.2d 591. As the insurer in Cruz conceded that it had been provided with the requisite information without the EUO, we held that the contract provision requiring an EUO was unenforceable on Cruz ’s facts. Id. at 590 n. 1, 600-601, 648 N.W.2d 591.
Cruz ’s analysis offers critical insight into the nature of what the no-fault law comprehensively regulates. It described the no-fault system as "a comprehensive legislative enactment designed to regulate the insurance of motor vehicles in this state and the payment of benefits resulting from accidents involving those motor vehicles. " Id. at 595, 648 N.W.2d 591 (emphasis added). When Citizens Ins. Co. , State Farm , Kneeland , and Cruz are read together, it becomes apparent that the comprehensive nature of the Insurance Code's regulation of no-fault insurance functions to ensure that there is applicable insurance for accidents and that benefits get paid. Citizens and State Farm both struck down agreements that purported to rearrange which insurer had to pay benefits, while Cruz struck down a policy provision that interfered with the payment of benefits. State Farm also noted that agreements that purport to rearrange which insurer is supposed to pay benefits also run the risk of leaving no insurer available to pay benefits. Meanwhile, Kneeland upheld an agreement that did not relate to the payment of mandatory benefits. The indemnification agreement at issue does not implicate the Cruz concerns. There is no dispute that the bus was "insured" (inasmuch as the city had satisfied the Secretary of State it could self-insure under MCL 500.3101(5) ), and there is no dispute that the benefits required by statute to be paid to Bronner and his caregivers were paid. Cruz clearly acknowledges that the Insurance Code's silence about a particular contractual provision may pose interpretive challenges in the right circumstances; but to implicate Cruz ’s concern, the contractual provision must, at minimum, relate to the insurance of motor vehicles or the payment of benefits resulting from motor vehicle accidents. This agreement implicates neither, but rather requires a vendor to reimburse the insurer for the cost of benefits compensating for an injury caused by the vendor's negligence. Where, as here, the agreement does not jeopardize the availability of applicable insurance or the payment of mandatory benefits, it falls outside our anti-shifting rule. As a result, no improper shift of liability as contemplated by Kneeland is implicated in this case, because a vendor reimbursing the insurer for the cost of mandatory benefits the vendor caused the insurer to pay out does not relate to either the availability of insurance or the payment of benefits.
It is not clear in hindsight why we performed the textual analysis of the meaning of "damages" in Kneeland to begin with. Even if the word "damages" could have been understood to include mandatory no-fault benefits, all that was at issue in Kneeland were nonmandatory collision benefits. Even if the Kneeland agreement had expressly stated that the borrower of the vehicle was accepting liability for both mandatory no-fault benefits and other nonmandatory damages, it is difficult to imagine we would have disallowed recovery of the nonmandatory damages simply because the agreement improperly shifted liability for mandatory benefits. We would presumably have construed the contract "to harmonize [it] with the statute," Cruz , 466 Mich. at 599, 648 N.W.2d 591, and enforced it to the extent that it was enforceable, but no further. This counsels further against overreliance on Kneeland .
The Court of Appeals similarly misconstrued the portions of the Insurance Code allowing no-fault insurers to seek reimbursement for payment of some benefits as implicitly excluding any other reimbursement mechanism (such as the indemnification provision at issue). It identified the Michigan Catastrophic Claims Association (the MCCA), MCL 500.3104, the Michigan Assigned Claims Plan (the MACP), MCL 500.3171, and the ability for no-fault insurers to impose a lien on tort damages recovered by some no-fault beneficiaries, MCL 500.3116, as the stated reimbursement opportunities for no-fault insurers under the Insurance Code. In other words, it effectively relied on the negative-implication canon, expressio unius est exclusio alterius , that in stating some options, other options must not exist. However, this "doctrine properly applies only when the unius (or technically, unum , the thing specified) can reasonably be thought to be an expression of all that shares in the grant or prohibition involved." Scalia & Garner, Reading Law: The Interpretation of Legal Texts (St. Paul: Thomson/West, 2012), p. 107. "Common sense often suggests when this is or is not so," id. , and this is such a case: we do not believe these options can be construed as "an expression of all that shares in the grant" of avenues for reimbursement. Rather, each of them responds to specific problems unrelated to the issue presented.
Expressio unius est exclusio alterius means "[e]xpress mention in a statute of one thing implies the exclusion of other similar things." Detroit v. Redford Twp. , 253 Mich. 453, 456, 235 N.W. 217 (1931).
First, the MCCA "was created by the Legislature in 1978 in response to concerns that Michigan's no-fault law ... placed too great a burden on insurers, particularly small insurers, in the event of ‘catastrophic’ injury claims." In re Certified Question , 433 Mich. 710, 714, 449 N.W.2d 660 (1989). "Its primary purpose is to indemnify member insurers for losses sustained as a result of the payment of personal protection insurance benefits beyond the ‘catastrophic’ level ...." Id. at 714-715, 449 N.W.2d 660. The MCCA spreads the cost of these catastrophic claims across all no-fault insurers in Michigan to equalize competitive imbalances between larger and smaller insurers and make the amount of cash on hand needed more predictable for insurers. See id. at 714 n. 2, 449 N.W.2d 660. Rather than being a substitute for reimbursement, it is, in effect, an entitlement for insurers—a cumulative remedy they enjoy above and beyond any other opportunities they may have to recoup the cost of benefits paid.
Indeed, if the MCCA is merely "a set security meant to assist against certain circumstances," which is to say, "when the PIP amount contracted by the insurer exceeds the statutory threshold," United States Fidelity Ins. & Guaranty Co. v. Mich. Catastrophic Claims Ass'n (On Rehearing) , 484 Mich. 1, 17-18, 795 N.W.2d 101 (2009), then the extent of the MCCA's obligation to its members may well be informed by the extent to which those members might be able to recoup such costs. If the MCCA "shall provide ... indemnification for 100% of the amount of ultimate loss sustained under [PIP] coverages in excess of" $580,000, MCL 500.3104(2)(o), then the degree to which insurers can be indemnified for their PIP losses before looking to the MCCA may affect the size of the "ultimate loss sustained."
Second, the MACP is a benefit to persons injured in motor vehicle accidents who otherwise do not have applicable insurance benefits. It imposes, by statute, the obligation of providing no-fault benefits to persons injured in motor vehicle accidents if an applicable no-fault policy cannot be identified, MCL 500.3172(1), on all no-fault insurers licensed to do business in Michigan. In other words, the MACP obliges them to function as insurers of last resort even as to some injured persons with whom the insurer does not have an existing insurance relationship, making "insurance companies ... the instruments through which the Legislature carries out a scheme of general welfare." Shavers v. Attorney General , 402 Mich. 554, 597, 267 N.W.2d 72 (1978). That the Insurance Code creates a mechanism in MCL 500.3385 by which insurers may pass on to their customers the cost of benefits the insurers must pay out by statutory fiat does not derogate from the insurers’ prerogative at common law to freely enter into contracts with vendors which may include indemnification provisions. Finally, the limited opportunity under MCL 500.3116 to recover certain benefits paid out does not imply the inability of an insurer to reach an indemnity agreement with a vendor. The statute allows "personal injury protection no-fault benefits [to] be reduced to the extent the insured has received equivalent compensation from tort judgments arising from accidents outside of the state, from accidents with uninsured motorists, and from intentionally caused harm." Tebo v. Havlik , 418 Mich. 350, 367, 343 N.W.2d 181 (1984). In such cases, the insurer is reducing its liability to (or recovering from) its beneficiaries. Section 3116 is, in effect, an exception that proves a rule: by providing a limited avenue by which a no-fault insurer can offset its liability to its own beneficiary, it implicitly denies other options at recovering from a beneficiary and confirms the no-fault system's focus on the relationship between insurers and their insureds and beneficiaries—not the relationship between insurers and their vendors. When we upheld the theoretical viability of EUOs in no-fault policies, we observed that "[t]he discovery tools provided in the [no-fault law] are not comprehensive" and rejected the argument that "the provision of some discovery tools by the act—tools that address limited aspects of the insurer's postclaim information needs—precludes the parties from contracting for the use of other discovery tools including those such as EUOs that enable insurers to directly gather information from the insured." Cruz , 466 Mich. at 598 n. 14, 648 N.W.2d 591. Much the same can be said about the no-fault law's reimbursement options. They are not comprehensive, and the fact that they are offered does not preclude the parties from contracting for other reimbursement methods. This is all the more apparent when the indemnification agreement at issue does not relate to "the insurance of motor vehicles in this state [or] the payment of benefits resulting from accidents involving those motor vehicles." Id. at 595, 648 N.W.2d 591. It does not alter the relationship between the insurer and its insured or its beneficiaries, and it does not transform the nature of benefits paid by the insurer to its beneficiaries into something else. It therefore does not conflict with the Insurance Code. IV. CONCLUSION
Section 3116 may also address GFL's concerns that a ruling in the city's favor here could validate other cost-recovery agreements that might be offensive to the no-fault system. A reimbursement clause that effectively changed an insurer's relationship with its insureds or beneficiaries —such as an agreement that the insurer would pay out benefits but asserted a right to subsequent reimbursement from the beneficiary—would presumably fall within the comprehensive scope of the statute and not be permitted. Transforming insurance benefits into the functional equivalent of a loan would change the character of the payments being made. By allowing a limited ability to claw back benefits from a beneficiary, § 3116 could certainly be read as implicitly precluding other such arrangements.
GFL argues that the city should not be treated any differently than more traditional no-fault insurers. We agree. If an ordinary insurance company reached an agreement with the vendor it hired to plow its parking lot in the winter that the plower would reimburse the insurer for accidents caused by the plower's negligence, such an agreement would be enforceable under today's ruling. That the city has far more opportunities to reach such agreements—and traditional insurers far fewer—is presumably offset by the fact that insurers are in the business of issuing no-fault insurance, while the city is in the business of providing a full panoply of municipal services and self-insures incidentally to that role. Regardless of the differing opportunities for an insurer to reach an indemnification agreement with a vendor, we conclude that such agreements are enforceable, and the contrary decision of the Court of Appeals is reversed.
McCormack, C.J., and Zahra, Bernstein, Cavanagh, and Welch, JJ., concurred with Clement, J.
Viviano, J. (concurring).
I agree with the result reached by the majority for many but not all of the reasons given in its opinion. I write separately to again highlight one larger issue that has escaped sustained attention in this area of the law: the appropriate analytical framework for addressing the vendor's claim that the no-fault act precludes enforcement of the contractual indemnity provision at issue. See Meemic Ins. Co. v. Fortson , 506 Mich. 287, 300-301 n. 7, 954 N.W.2d 115 (2020) (noting the unsettled state of the interpretive framework in this area). Whatever approach we may decide to adopt in a future case, I believe the majority's approach here relies too heavily on ascertaining the statute's broad purposes.
The core issue, as the majority states, is whether the parties’ contractual indemnity agreement is unenforceable because it violates public policy as represented by the no-fault act. In our most recent opinion addressing this general topic, we observed that our caselaw contains various standards for determining whether the no-fault act, or various provisions of it, has abrogated the common law and thereby precludes the parties from incorporating certain common-law defenses in their insurance contracts. Id. at 300-301 n. 7, 954 N.W.2d 115. Some of our cases hold that the intent to abrogate must be clearly stated in the statute, whereas other cases indicate that the comprehensiveness of the statutory scheme can indicate abrogation. Id.
The majority today opts for the latter standard, which is how the case was argued and decided below, although no one—including the majority—has expressly examined whether this is the appropriate interpretive methodology for assessing this issue. In adopting this approach, the majority's framework resembles a field-preemption analysis by asking whether the no-fault act has impliedly preempted parties from contracting for indemnification. See generally Mich. Gun Owners, Inc. v. Ann Arbor Pub. Sch. , 502 Mich. 695, 702-708, 918 N.W.2d 756 (2018) (discussing preemption in general and field preemption specifically). Some support exists for this approach. For example, we have often used the term "preemption" when discussing abrogation. See, e.g., Kyser v. Kasson Twp. , 486 Mich. 514, 539, 786 N.W.2d 543 (2010). More directly, the Delaware Supreme Court has held that because these concepts are so similar, a preemption-like analysis is applicable to resolve questions of abrogation. See A.W. Fin. Servs., S.A. v. Empire Resources, Inc. , 981 A.2d 1114, 1122 (Del., 2009) ("Although preemption and superseder [i.e., common-law abrogation] are analytically distinct concepts, they both involve the same inquiry: has one body of law replaced another? For that reason, the preemption analytical framework is a useful tool to conduct our analysis of whether the Escheat Statute has superseded common law claims.").
To the extent that a field-preemption analysis applies here—and I would take the opportunity in a future case to more closely analyze this question—my only significant disagreement with the majority is how it applied that analysis. It is difficult to determine when a field has been impliedly preempted by a statute. Scalia & Garner, Reading Law: The Interpretation of Legal Texts (St. Paul: Thomson/West, 2012), § 47 (discussing the presumption against federal preemption of state law). At bottom, field preemption "is really ‘a species of conflict preemption,’ " in that it is triggered when a legal provision trenches upon (i.e., conflicts with) a statute's occupation of a field. Id. , p. 290, quoting English v. Gen. Electric Co. , 496 U.S. 72, 79 n. 5, 110 S. Ct. 2270, 110 L. Ed. 2d 65 (1990). That a conflict lies at the heart of field preemption is important to keep in mind because it is very easy for the field-preemption analysis to "exalt extratextual purpose above statutory text." Note, Preemption as Purposivism's Last Refuge , 126 Harv. L. Rev. 1056, 1057 (2013). The reason is that "field preemption essentially impl[ies] additional statutory clauses beyond the statute's text, clauses that mandate preemption." Id. at 1064. In addition, "choosing the correct field definition" is difficult and critical because "[d]efining the field at a certain level of generality becomes the entire game." Id. at 1067. As a result, I believe it is important to stick to the text as much as possible when defining the field. One way to do this is by recognizing that a statute's occupation of one area of the law does not necessarily mean that it occupies adjacent areas as well. Cf. A.W. Fin. Servs. , 981 A.2d at 1124 ("With one exception, the Escheat Statute does not impliedly supersede other areas of the common law, because there is no ‘fair repugnance’ between the statute and common law areas that are not related to escheat.").
In this case, by investigating the no-fault act's few scattered provisions concerning reimbursement, the majority thoroughly demonstrates that the act does not occupy this area of law. See ante at 317–19. And through the same analysis of these specific statutory provisions, the majority ably explains why the indemnity agreement at issue does not directly conflict with the operation of any other provision in the no-fault act. This analysis, in my view, is generally sufficient to dispose of the case. It shows that the no-fault act does not occupy the field of indemnification and that none of the handful of relevant provisions conflicts with the indemnification contract at issue.
Implied conflict preemption is another type of preemption under which a provision directly conflicts with state law, i.e., when the provision permits what the statute prohibits or vice versa. DeRuiter v. Byron Twp. , 505 Mich. 130, 140, 949 N.W.2d 91 (2020). As with the field-preemption inquiry, no one here expressly framed the case in these terms. But the mode of analysis used here, in searching for a conflict, is similar, and I would also consider, in an appropriate case, whether this framing is helpful.
I therefore cannot agree that the majority's assessments of the sweeping scope and purpose of the no-fault scheme have much, if any, analytical significance. That is, I cannot agree that the enforceability of the indemnification contract at issue turns upon whether a court considers it to be consistent with the broadly characterized statutory goals of regulating the insurance of motor vehicles and requiring payment of benefits. See ante at 316–17. I do not believe that the proper question in cases like the present one is whether a contract provision "implicate[s]" or "relate[s] to" either of these broadly defined purposes of the no-fault schemes. Ante at 316–17. Rather, the case calls for a closer examination of the statutory text, such as the majority itself provides in addition to its assessments of the statute's broader objectives.
The majority's use of statutory purpose here is problematic in at least two respects. One is that it further aggrandizes the purposes we have attributed (without much assessment of the text) to the no-fault act. In Shavers v. Attorney General , 402 Mich. 554, 578-579, 267 N.W.2d 72 (1978), we articulated a somewhat narrower purpose of the statute as "provid[ing] victims of motor vehicle accidents [with] assured, adequate, and prompt reparation for certain economic losses" through the means of compulsory insurance. We subtly expanded this in Cruz v. State Farm Mut. Auto. Ins. Co. , 466 Mich. 588, 595, 648 N.W.2d 591 (2002), suggesting that both the "insurance of motor vehicles ... and the payment of benefits" were purposes of the act. Today, the majority gives these purposes teeth, holding that contract provisions that "relate" to or "implicate" these broad purposes can thereby be rendered unenforceable. Cruz did not establish such an aggressive use of these statutory purposes. Rather, it stated that contracts that "contravene[ ] the requirements of the no-fault act by imposing some greater obligation upon one or another of the parties [are], to that extent, invalid." Id. at 598, 648 N.W.2d 591. That inquiry involves a more careful examination of the statutory provisions at issue.
A second troubling aspect of the majority opinion is the imprecise words it uses to describe when the purposes of the no-fault act preclude enforcement of a contract: if the provision "implicates" or "relates to" the purposes. It is uncertain how these criteria will be met, as it seems likely that many provisions in an insurance contract will "implicate" or "relate to" either insuring motor vehicles or paying benefits. The majority appears to give these terms a limited scope, implying that a provision implicates or relates to a purpose only if it results in denying insurance or benefits owed under the act. But if that is so, why does the majority define the purposes so broadly? There are specific statutory sections that relate to insuring vehicles and paying benefits. Under Cruz , a court should examine those particular sections to determine whether they are contravened by the contractual provision at issue. By generalizing the purposes of the no-fault act, the majority today suggests that contractual agreements are in jeopardy even if they do not violate a particular provision but instead have some connection with a broadly conceived statutory purpose. See Preemption as Purposivism's Last Refuge , 126 Harv. L. Rev. at 1067.
As I said at the outset, I agree with the conclusion the majority reaches and with much of its work in getting there. I agree that the no-fault act is not a comprehensive regulation of indemnification agreements and that none of the pertinent statutory provisions conflicts with the agreement here. Therefore, I agree that the indemnification contract does not violate the no-fault act and should be upheld. I do not believe, however, that to reach this conclusion we should rely on the statute's abstract goals as defined by this Court. While the proper interpretive framework remains somewhat unclear—in particular, whether preemption principles can illuminate the interpretation of the statute—I cannot subscribe to a methodology that relies so heavily on statutory purposes. For these reasons, I respectfully concur.
The majority also extends its holding to insurers, even though the city here is a self-insurer. See ante at 318–20. While the majority's conclusion might very well be correct, the majority has not offered any supporting analysis and, in any event, it is unnecessary to reach this issue. Consequently, I do not join this portion of the majority opinion.