Section 3 Malfunction Theory Limited to Lost/Destroyed Products: No Inference of Defect When the Product Remains Available for Inspection

Case: Millentine Coates v. Ford Motor Company (3d Cir. Aug. 31, 2026)
Court: United States Court of Appeals for the Third Circuit
Posture: Appeal from summary judgment for Ford on a Virgin Islands strict products-liability claim

1. Introduction

Millentine Coates alleged she was injured when the airbags in her 2002 Ford Explorer “spontaneously deployed” while she was driving on St. Thomas, causing disorientation and a crash into a hillside. She sued Ford Motor Company in strict products liability, but could not identify a specific defect in the restraint system or its restraint control module (“RCM”). Instead, she sought to proceed under the “malfunction theory,” a circumstantial-evidence doctrine reflected in Section 3 of the Restatement (Third) of Torts: Products Liability (“Section 3”).

The key issue was whether Coates could invoke Section 3’s relaxed inference of defect when the allegedly defective product (the vehicle/airbag system/RCM) was not lost or destroyed and was inspected—though not exhaustively, despite recommendations for further testing. A second issue was evidentiary sufficiency: even if Section 3 applied, did the record support a reasonable inference that a defect existed at the time the vehicle left Ford’s control, rather than arising from years of use, prior accidents, repairs, or modifications?

2. Summary of the Opinion

The Third Circuit affirmed summary judgment for Ford. It held:

  • The Supreme Court of the Virgin Islands has “all but adopted” Section 3, and the Third Circuit predicted it would adopt Section 3 formally.
  • Nevertheless, Section 3’s malfunction theory is unavailable where the allegedly defective product is available for inspection and the plaintiff remains “theoretically capable” of proving a specific defect through direct evidence. Missing RCM crash/deployment data was not the “functional equivalent of destruction,” particularly because additional inspection/testing was recommended but not performed.
  • Even assuming Section 3 applied, Coates’s evidence was too speculative to show (by a preponderance) that the most likely cause was a defect existing at the time of sale—especially given 14 years of use, at least 93,000+ miles, multiple owners, prior accidents, repairs, and modifications (including unreconnected seatbelt pretensioners).

3. Analysis

3.1. Precedents Cited (and How They Shaped the Decision)

A. Virgin Islands adoption framework for Restatement-based products liability

  • Banks v. Int'l Rental & Leasing Corp., 55 V.I. 967 (V.I. 2011): Adopted Restatement (Third) of Torts: Products Liability § 1, grounding strict products-liability doctrine in Virgin Islands law and framing the baseline rule that sellers/distributors are strictly liable for harm caused by defective products.
  • Davis v. UHP Projects, Inc., 74 V.I. 525 (2021): Adopted Section 2 and stated that the “same analysis” would support adopting Section 3. The Third Circuit relied on this explicit signal to “predict” formal adoption of Section 3, while still controlling the scope of Section 3’s availability.
  • Tutein v. Ford Motor Co., 67 V.I. 34 (V.I. Super. Ct. 2016): The principal Virgin Islands malfunction-theory decision, involving an airbag system “prematurely discarded.” The Third Circuit used Tutein to (i) illustrate Section 3’s typical use-case (product unavailable through no fault of plaintiff), and (ii) show that local VI precedent did not resolve whether Section 3 applies when the product remains available.

B. Summary judgment standards and limits on inference

  • Wiest v. Tyco Elecs. Corp., 812 F.3d 319 (3d Cir. 2016); SodexoMAGIC, LLC v. Drexel Univ., 24 F.4th 183 (3d Cir. 2022); Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986); D.E. v. Cent. Dauphin Sch. Dist., 765 F.3d 260 (3d Cir. 2014): These cases supplied the framework that courts must draw reasonable inferences for the nonmovant but must reject speculation, conjecture, and suspicion as insufficient to create a genuine dispute. That theme directly drove the alternative holding: even under Section 3, Coates’s proof did not lift the claim above speculation.

C. Erie/choice-of-law posture for Virgin Islands claims in federal court

  • Edwards v. HOVENSA, LLC, 497 F.3d 355 (3d Cir. 2007): Confirmed Erie’s application to the District Court of the Virgin Islands and articulated the “predict how the Supreme Court of the Virgin Islands would decide” methodology. This underwrote the Third Circuit’s prediction that the VI Supreme Court would adopt Section 3, while still defining its contours.

D. The malfunction theory’s content, rationale, and boundaries

  • Metro. Prop. & Cas. Ins. Co. v. Deere & Co., 25 A.3d 571 (Conn. 2011): A key doctrinal guidepost cited for (i) the malfunction theory’s narrow role (does not relieve proof of all elements), (ii) the increased speculation risk because products typically leave defendant’s control, and (iii) the two-inference structure (defect at incident; defect at time of sale). The Third Circuit borrowed Metro. Prop. repeatedly to justify limiting Section 3 to situations where direct proof is genuinely unavailable.
  • Roskop Dairy, L.L.C. v. GEA Farm Techs., Inc., 871 N.W.2d 776 (Neb. 2015): Reinforced that malfunction theory is not a substitute for proving proximate cause or damages and criticized use of the doctrine when the product is still available. This supported the court’s decision to reject Section 3 where Coates had access to the RCM and could have conducted further inspections.
  • Barnish v. KWI Bldg. Co., 980 A.2d 535 (Pa. 2009): Used for the proposition that plaintiffs often must negate other potential causes and that Section 3 does not eliminate the plaintiff’s burden; it also helped frame the “most likely cause” requirement.
  • Myrlak v. Port Auth. of N.Y. & N.J., 723 A.2d 45 (N.J. 1999): Provided the analytic distinction that malfunction theory, unlike res ipsa loquitur, permits two inferences—defect and defect-at-sale—thus warranting careful policing of when the doctrine is available.
  • Dansak v. Cameron Coca-Cola Bottling Co., 703 A.2d 489 (Pa. Super. Ct. 1997); DeWitt v. Eveready Battery Co., 565 S.E.2d 140 (N.C. 2002): Listed the kinds of circumstantial evidence commonly used to support a malfunction inference (malfunction itself, expert testimony, timing, similar accidents, elimination of other causes, etc.). The court acknowledged Coates had some “malfunction” evidence but found it insufficient on the crucial “defect-at-sale” question.

E. Res ipsa loquitur origins and the “access to evidence” justification

  • Byrne v. Boadle, 2 H. & C. 722, 159 Eng. Rep. 299 (Ex. 1863): Used as the seminal res ipsa case to emphasize why inference doctrines exist—plaintiffs lack access to direct evidence “through no fault of her own.”
  • Coastal Air Transp. v. Royer, 64 V.I. 645 (V.I. 2016): Cited for res ipsa’s “extraordinary event under the defendant’s control,” underscoring the control-based logic that is weaker in products cases and thus increases the need to cabin malfunction theory.

F. Decisions limiting malfunction theory when the product is available

  • Decato v. Brandfon Motors, Inc., 2013 WL 4873069 (Conn. Super. Ct. Aug. 20, 2013); Ellis v. Beemiller, Inc., 910 F. Supp. 2d 768 (W.D. Pa. 2012): These authorities were cited for the view that malfunction theory is properly limited to cases where the product/components are destroyed, unavailable, or inaccessible. They anchored the Third Circuit’s central holding: the doctrine’s rationale collapses when direct inspection remains feasible.

G. “Defect at time of sale” and prolonged use/alteration as inference-killers

  • Living & Learning Ctr., Inc. v. Griese Custom Signs, Inc., 491 A.2d 433 (Conn. App. Ct. 1985): Quoted for the critical analytical separation between (i) inferring a defect at the time of the incident and (ii) inferring that defect existed at sale.
  • Hardin v. Montgomery Elevator Co., 435 So. 2d 331 (Fla. Dist. Ct. App. 1983): Provided the multi-factor list (age, repairs, misuse, maintenance history) for judging whether a plaintiff has made a prima facie showing that a defect existed within the manufacturer’s control.
  • Kuisis v. Baldwin-Lima-Hamilton Corp., 319 A.2d 914 (Pa. 1974); Pryor v. Lee C. Moore Corp., 262 F.2d 673 (10th Cir. 1958); Scanlon v. Gen. Motors Corp., Chevrolet Motor Div., 326 A.2d 673 (N.J. 1974): Used to emphasize that prolonged use substantially weakens the inference that a defect existed at sale; “the older a product is, the more difficult” the proof becomes.
  • Sochanski v. Sears, Roebuck and Co., 621 F.2d 67 (3d Cir. 1980): Recognized that origin questions are often for the jury, but only where “reasonable and well-balanced minds” could find the defect existed when delivered—language the Third Circuit contrasted with the lopsided record here.
  • Ford Motor Co. v. Ridgway, 135 S.W.3d 598 (Tex. 2004): Cited as an example where product age, mileage, multiple owners, and repairs defeated an indeterminate-defect inference—paralleling the Third Circuit’s view that Coates’s record left “equally likely” non-manufacturer causes.

H. Procedural forfeiture

  • In re Wettach, 811 F.3d 99 (3d Cir. 2016): Used to hold Coates forfeited any developed explanation for not pursuing additional RCM teardown/frontal sensor testing, reinforcing the court’s skepticism toward invoking Section 3 to fill an evidentiary gap left by litigation choices.

3.2. Legal Reasoning

The court’s reasoning proceeds in two decisive steps.

Step 1: Section 3 is (likely) part of Virgin Islands law—but its “relaxed” inference is not universally available

The court first predicted the Supreme Court of the Virgin Islands would adopt Section 3, leaning heavily on Davis v. UHP Projects, Inc. and the Restatement-centered structure already adopted in Banks v. Int'l Rental & Leasing Corp.. But adoption did not resolve scope. The Third Circuit treated Section 3 as a narrow, rationale-driven evidentiary rule analogous to res ipsa loquitur: it exists to prevent a plaintiff from losing solely because direct proof is unavailable “through no fault of her own.”

From that premise, the court articulated the new limiting principle: when the product is available and the plaintiff is “theoretically capable” of proving a specific defect through inspection/testing, Section 3 does not “level the playing field” and should not be used to bridge a gap that could be filled by diligent examination. Missing RCM logs did not transform an available module/system into a “destroyed” product; other direct-evidence avenues remained open (RCM teardown; frontal accelerometer location and off-line testing), and the plaintiff did not pursue them.

Step 2: Even if Section 3 applied, the record could not support the key inference that the defect existed at sale

Section 3 requires evidence that (a) the incident is of a kind that ordinarily occurs due to defect, and (b) it was not solely the result of other causes besides a defect existing at the time of sale/distribution. The court credited Coates with some circumstantial evidence of malfunction (her testimony and contemporaneous report; the RCM’s failure to record). But the analysis turned on the second prong: whether a reasonable factfinder could conclude the “most likely” cause was a defect attributable to Ford at the time of sale.

The extensive alternative-cause landscape defeated that inference: 14 years of use, 93,000+ miles as of 2009, multiple owners, prior accidents, engine replacement, wiring repair, later collision repairs, and a seat replacement with unreconnected pretensioners. Against that, Coates offered only that the RCM was original, the vehicle seemed in good condition, and the historic diagnostic codes were “suspicious”—yet her expert could not explain the codes’ meaning or link them (to a reasonable degree of engineering certainty) to spontaneous deployment. The court deemed the defect-at-sale conclusion speculative.

3.3. Impact

  • Doctrinal gatekeeping in the Virgin Islands (and persuasive authority elsewhere): The opinion supplies a clear operational limitation on Section 3 in Virgin Islands diversity cases: the malfunction theory is generally reserved for when the product (or key component) is lost, destroyed, or otherwise genuinely unavailable/inaccessible for inspection, not merely when evidence is incomplete.
  • Litigation incentives and discovery strategy: Plaintiffs pursuing indeterminate-defect theories will face strong pressure to exhaust feasible inspections and testing. A recommended teardown or component test left undone may undermine access to Section 3 and, independently, strengthen summary judgment arguments that the claim rests on conjecture.
  • Heightened importance of “defect-at-sale” proof for older products: The court’s prolonged-use analysis signals that, for mature products with long service lives, multiple owners, and substantial repair history, plaintiffs must develop affirmative evidence excluding age/repair/alteration causes—not simply prove “something went wrong.”
  • Expert testimony limits: The decision illustrates that “suspicious” diagnostic artifacts, without interpretive linkage and causal explanation, may be insufficient both to establish defect and to anchor the defect temporally to the point of sale.

4. Complex Concepts Simplified

  • Strict products liability (Restatement (Third) §§ 1–2): Liability can attach without proving negligence if the product was defective (manufacturing or design) when sold and caused harm.
  • “Malfunction theory” / Section 3: A circumstantial-evidence shortcut allowing a jury to infer a defect (and that it existed at sale) without identifying a specific broken part—typically used when the product is destroyed or unavailable.
  • Res ipsa loquitur: A negligence doctrine allowing inference of negligence when the accident is the kind that ordinarily does not occur without negligence and the instrumentality was in the defendant’s control; Section 3 is analogous but can be more speculative because products usually leave the manufacturer’s control.
  • Summary judgment (Rule 56): The court ends the case before trial if no reasonable jury could find for the nonmoving party based on admissible evidence; speculation cannot defeat it.
  • RCM (restraint control module): The vehicle’s airbag/seatbelt “computer” that senses crashes, commands deployments, and may record event data. Here, the RCM did not record a crash/deployment event, but its recording function was described as independent of deployment functions.
  • “Defect existing at the time of sale”: Even if a product failed later, the plaintiff must connect the defect to when it left the manufacturer, not to later wear, repairs, collisions, or modifications.

5. Conclusion

Millentine Coates v. Ford Motor Company establishes a consequential limitation on Section 3’s malfunction theory in Virgin Islands products-liability practice: the doctrine’s relaxed inference is generally unavailable when the allegedly defective product remains available for inspection and the plaintiff could, in principle, prove a specific defect through direct testing. The court also underscores that, even where malfunction evidence exists, plaintiffs must present non-speculative proof that the most likely cause was a defect attributable to the manufacturer at the time of sale—an especially demanding showing for older, heavily used, and repeatedly repaired products.