Union Pacific Railroad Company v. Plano Molding Company

In the United States Court of Appeals For the Seventh Circuit ____________________ Nos. 14-1171 & 14-1189 KAWASAKI KISEN KAISHA, LTD., “K” LINE AMERICA, INC., and UNION PACIFIC RAILROAD CO., Plaintiffs-Appellants, v. PLANO MOLDING CO., Defendant-Appellee. ____________________ Appeals from the United States District Court for the Northern District of Illinois, Eastern Division. No. 07 C 5675 — Harry D. Leinenweber, Judge. ____________________ ARGUED JANUARY 6, 2015 — DECIDED MARCH 31, 2015 ____________________ Before FLAUM, WILLIAMS, and TINDER, Circuit Judges. FLAUM, Circuit Judge. On April 21, 2005, a Union Pacific freight train derailed in Oklahoma. The train was carrying two steel injection molds being delivered to Plano Molding Company in Illinois. The derailment occurred after the molds broke through the floor of their shipping container, causing that train car and many behind it to derail and re- sulting in approximately $4 million in total damage. The 2 Nos. 14-1171 & 14-1189 molds had been manufactured in China and shipped to the United States before being transferred to the Union Pacific train for the final leg of their journey. At issue in this case is the reason that the molds broke through the floor. The appellants are three companies that were involved in the shipment of the molds and sustained losses from the accident: Kawasaki Kisen Kaisha, Ltd., “K” Line America, Inc. (collectively, “K-Line”), and Union Pacific Railroad Co. (“Union Pacific”). They sued Plano, claiming that Plano was at fault because a company it hired packed the molds into the shipping container improperly, causing the floor of the container to break and ultimately causing the derailment. If true, Plano would be liable to appellants for breach of a warranty found in a document known as the “World Bill of Lading,” which provided contractual terms for the shipment of the molds. Plano, in defense, argued that the molds were properly packed and that they fell through the floor of the container because the container was defec- tive. In a bench trial, the district court found in favor of Plano, concluding that appellants had not proved that the molds were improperly packed. The court also held that the de- railment was in fact caused by deficiencies in the container. On appeal, the plaintiffs contest these factual conclusions, as well as a number of other aspects of the district court’s opin- ion. For the reasons set forth below, we affirm. I. Background Plano is an Illinois corporation that designs, manufac- tures, and sells plastic storage boxes. In 2004, Plano identi- fied a need to purchase two new steel injection molds, which Nos. 14-1171 & 14-1189 3 it uses to manufacture its plastic boxes, and so it contacted CMT International, Inc. (“CMT”), a company that assists customers in the United States who wish to purchase prod- ucts from Asia. CMT provided Plano with bids, and Plano selected Kunshan Yuanjin Plastic & Electronic Co., Ltd. (“Kunshan”), a Chinese company, to make the molds. Kun- shan manufactured the molds and loaded them into wooden crates. Plano hired World Commerce Services (“World”), 1 a non- vessel operating common carrier, 2 to arrange for the ship- ment of the molds from China to Plano’s factory in Illinois. World then contracted with the THI Group LTD (“THI”) and K-Line for the physical shipment of the molds. THI, in turn, hired Shanghai Haixing Yuancang Container Transportation Co. (“Haixing”) to load the molds into an intermodal ship- ping container and hired Shanghai Ocean Tally Company as a “checker,” which ensured that the molds were properly loaded. That shipping container carried Plano’s molds (in- side of their wooden crates) from the moment they were packed in China until the Union Pacific train derailed in Ok- lahoma. Both World and K-Line issued bills of lading cover- ing the shipment of the molds. K-Line handled the ocean 1Plano contends that it was actually CMT that retained World’s services. For the purposes of this appeal, however, this dispute is irrelevant. 2 As their name suggests, non-vessel operating common carriers do not own or operate ships or other means of freight transportation. Rather, they act primarily as coordinators of shipping logistics, and contract with carriers to actually move goods. They do, however, assume common car- rier responsibility, 46 U.S.C. § 40102(6), (16), and often issue their own bills of lading. See, e.g., Sompo Japan Ins. Co. of Am. v. Norfolk S. Ry. Co., 762 F.3d 165, 168–69 (2d Cir. 2014). 4 Nos. 14-1171 & 14-1189 part of the voyage, but subcontracted the overland move- ment of the molds within the United States to Union Pacific. Once the shipping container carrying Plano’s molds arrived in the U.S., it was transferred from K-Line to Union Pacific, which began transporting the container by train from Cali- fornia to Illinois. During the voyage, the molds somehow broke through the floor of the shipping container, causing the train to derail in Oklahoma. The accident caused approx- imately $2 million in damage to the cargo of K-Line’s cus- tomers and $2 million in costs to Union Pacific. Appellants sued in federal court, attempting to hold Pla- no liable for certain damages caused by the derailment and seeking indemnification for claims made against appellants by third parties who suffered damages in the accident. The district court initially granted summary judgment in Plano’s favor on all of appellants’ claims, but this court reversed on one claim—a breach of contract claim based on a warranty in the bill of lading issued by World. Kawasaki Kisen Kaisha, Ltd. v. Plano Molding Co., 696 F.3d 647, 657–58 (7th Cir. 2012). We remanded to the district court to determine whether Plano was subject to the terms of the World Bill of Lading. After a one-day trial, the district court held that Plano was bound by the World Bill of Lading and could be held liable to appel- lants if it violated the terms of that agreement. Kawasaki Kisen Kaisha, Ltd. v. Plano Molding Co., No. 07 C 5675, 2013 WL 3791609, at *8 (N.D. Ill. July 19, 2013) (“Kawasaki I”). Pla- no does not appeal that ruling. The district court then held a three-day trial to determine whether Plano breached Clause 10(2) of the World Bill of Lading, which states: Nos. 14-1171 & 14-1189 5 If Carrier receives the goods already packed in- to containers: … (2) Merchant warrants that the stowage and seals of the containers are safe and proper and suitable for handling and carriage and indem- nifies Carrier for any injury, loss, or damage caused by the breach of this warranty Plano, the parties agree, was a “Merchant” as defined in Clause 2(3) of the World Bill of Lading. The World Bill of Lading defined World as the “Carrier,” but Clause 3 pro- vides that World’s contractors and subcontractors— including all of the appellants—are entitled to all of the Car- rier’s rights under the Bill of Lading. The trial consisted largely of expert testimony. Appellants sought to prove that Plano breached Clause 10(2) by im- properly stowing and securing the molds in the shipping container. 3 There was no direct evidence of how the molds were packed into the container, and neither party presented witnesses who were involved in the loading of the container. That is in part because the container was packed in China, making it difficult to obtain this information. 3 As stated above, Plano itself did not actually pack its molds into the shipping containers; rather, Haixing did, and its work was checked by Shanghai Ocean Tally. However, assuming, as the district court did, that these companies were hired by Plano to pack the molds, Plano would be liable for those companies’ breach of Clause 10(2) just as if Plano itself had packed the container. Therefore, to avoid unnecessary complexity, we will conduct our analysis as if Plano itself packed the molds. 6 Nos. 14-1171 & 14-1189 The derailment itself also made it very difficult to deter- mine after the fact how the container had been loaded. Much of what was inside of the container before the crash ended up outside of it, spread over the miles-long crash site. About 60% of the floor of the container was missing after the acci- dent, allowing much to fall out during the derailment. There was, however, some physical evidence about the stowage of the molds. Together, the two molds weighed 25,000 pounds, and each was packed into its own wooden crate. One mold was much larger than the other—it weighed 18,900 pounds, while other weighed 6,100 pounds. When packing shipping containers, packers use wooden pallets, wooden beams, and other types of dunnage (all types of inexpensive waste mate- rials) to support and distribute the weight of cargo. After the derailment, all that was found in the shipping container was one 43-inch long wooden pallet, which presumably support- ed, at least in part, the weight of one of the crates. There were remnants of wooden pallets found at the accident site, as well as a wide variety of other debris that could have served as dunnage. Exhibit 281, a photograph taken at the derailment scene, appears to show two pallets that survived the accident, as well as several long wooden boards. It is not clear, however, which shipping container the pallets and wood came from, as the derailment caused multiple contain- ers to break open and spill their contents. Appellants’ theory of the case was that the molds were packed in a way that did not sufficiently distribute their weight in the container. They based this contention on the guidelines set forth in Circular 43-D, a publication of the As- sociation of American Railroads, an industry trade group. Appellants argued that the guidelines in Circular 43-D, which was entered into evidence, should serve as the stand- Nos. 14-1171 & 14-1189 7 ard of care in this case. Appellants did not claim that the to- tal weight of the molds violated the Circular 43-D guide- lines, but rather that the weight was too concentrated. Re- garding the concentration of weight in a shipping container, Circular 43-D states that “not more than 25,000 lbs. uniform- ly distributed in any 10 linear feet can be loaded.” Appellants’ argument was presented largely through its expert witness, Dr. Robert Vecchio. Vecchio theorized that the pallet found in the shipping container following the ac- cident was one of two identical 43-inch pallets that were originally used to support the molds, with each pallet sup- porting one crate. The larger of the two crates weighed 18,900 pounds; spread out over 43 inches, that translates to over 5,000 pounds per linear foot, which is double the amount allowed for in Circular 43-D. Vecchio admitted, however, that a 43-inch pallet would have been sufficient to disburse the weight of the smaller crate. Vecchio also argued that the crates were not properly se- cured (“lashed”) within the shipping container, which al- lowed the crates to exert “dynamic amplification”—in other words, bouncing—on the floor of the container. His theory was that the combination of a failure to secure the crates and a failure to distribute the crates’ weight properly led to over- stressing of the container floor, eventually causing the floor to break. To support his theory, Vecchio noted that no lash- ing material was found inside of the container or around the train following the derailment. Additionally, he referred to a number of “significant dents” in the wood floor of the con- tainer that were found after the derailment. According to Vecchio, these dents were caused by dynamic amplification 8 Nos. 14-1171 & 14-1189 of the crates, which only could have occurred if the crates were not properly lashed. The parties’ experts referred to numerous investigative reports about the derailment at trial. The most important re- port, on which both appellants and Plano relied heavily and which was introduced into evidence, was produced by a company called Intertek Caleb Brett. Appellants focused on the Intertek report’s statement that the crates were placed in the center of the container, arguing that this implied that the weight of the crates was not properly distributed. On the other hand, the Intertek report stated that the crates were se- cured and lashed, contrary to Vecchio’s theory. Another im- portant piece of evidence was a “Testification” produced by the Shanghai Ocean Shipping Tally Company, the company that supervised the loading of Plano’s molds into the ship- ping container. The Testification, like the Intertek report, stated that the cargo was loaded in the middle of the con- tainer. However, it also stated that the crates were “packed sound,” which could imply that the crates’ weight was properly distributed and that the cargo was properly lashed. Finally, a post-derailment email from Joana Feng of the THI Group to World’s John Wember stated that “wooden brack- ets” had been used “to fix the case[s] in order not to move when transmitting.” On cross-examination, Vecchio stated that he had not previously seen the Testification or Feng’s email, and therefore did not take either into account when formulating his report. The district court found that appellants had not provided sufficient evidence to prove that the weight of Plano’s crates was not distributed properly. Without any direct evidence as to how the container was packed, the court said that it was Nos. 14-1171 & 14-1189 9 pure speculation for Vecchio to assume that the crates were supported by two 43-inch pallets. Although only one pallet was found in the container after the derailment, the district court noted that a second pallet—which even Vecchio as- sumed had been present—could have been large enough to support the larger crate. Moreover, the crates may have been supported by other pieces of dunnage. According to the dis- trict court, any of these scenarios was equally plausible given that more than enough wood to adequately support the crates was found at the site of the accident. The district court also found that there was not enough evidence to support appellants’ claim that the crates had not been properly lashed. The court criticized Vecchio for disre- garding the Intertek report’s finding that the crates had been properly secured and lashed, when Vecchio otherwise relied heavily on the report and had said that it “provided the best available information” and that he “had no reason not to be- lieve it.” The district court also stated that Vecchio’s conclu- sion was belied by the email from Feng to Wember stating that the crates had been fixed to prevent movement. The dis- trict court downplayed the fact that no lashing material had been found at the accident site, stating that it could have been lost or destroyed during the accident, or that investiga- tors may have not even searched for it. The court also reject- ed Vecchio’s theory that the dents in the container’s floor were caused by dynamic amplification of the crates. The dents, the court noted, could have preexisted the loading of the crates into the container or could have been created when the crates were loaded or during the violent derail- ment. 10 Nos. 14-1171 & 14-1189 Both appellants and Plano also presented a great deal of evidence regarding the cause of the derailment, most of which focused on the condition of the shipping container. Because we ultimately agree with the district court’s conclu- sion that appellants failed to prove that Plano breached Clause 10(2), and therefore do not reach the issue of causa- tion, we provide only a brief summary of this evidence here. In essence, Plano and its experts argued that the container was defective, mostly due to the condition of its welds. Pla- no’s position was that the molds fell through the floor of the shipping container not because they were packed improper- ly, but rather because the floor of the container was defective and gave way. Appellants, conversely, argued that the welds, and the container as a whole, were sound. The district court agreed with Plano, finding that the shipping container was defective and that “defective welds in the shipping container caused the molds to fall through the bottom of the container and cause the derailment.” Kawasaki Kisen Kaisha, Ltd. v. Pla- no Molding Co., No. 07 C 5675, slip op. at 23 (N.D. Ill. Mar. 6, 2014) (“Kawasaki II”). II. Discussion Appellants’ central argument is that the district court clearly erred by finding that appellants failed to meet their burden of proving that Plano breached Clause 10(2), and that this breach caused the Union Pacific derailment. Before turning to this general issue, we address appellants’ argu- ments regarding more specific aspects of the district court’s opinion. We apply federal maritime law because jurisdiction exists under 28 U.S.C. § 1333. See Norfolk S. Ry. Co. v. Kirby, 543 U.S. 14, 24–25 (2004) (finding bills of lading involving Nos. 14-1171 & 14-1189 11 overseas shipment of goods to be maritime contracts even where the last leg of the journey was by rail). First, appellants argue that the district court ignored the guidance of Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993), when it assessed the testimony of their expert, Vecchio. Specifically, they claim that it was improper for the district court to find Vecchio’s theory less “credible” than the theories presented by Plano’s experts. Daubert, they argue, precludes a judge acting as a trier of fact from questioning an expert witness’s conclusions. Appellants misunderstand a court’s role in assessing expert testimony. It is true, as appel- lants point out, that the Daubert Court said a judge’s focus when assessing expert testimony “must be solely on princi- ples and methodology, not on the conclusions that they gen- erate.” Id. at 595. The Court, though, was referring to a judge’s analysis, under Federal Rule of Evidence 702, of whether to admit expert testimony 4 at all. Id. at 594–95. In making this determination, a judge acts as a “gatekeeper” to ensure that expert testimony is not admitted into evidence unless it is sufficiently reliable. Dhillon v. Crown Controls Corp., 269 F.3d 865, 869 (7th Cir. 2001). The district court here performed this gatekeeping analysis for each expert. In fact, the court did so both before the trial and in its final opinion, after both appellants and Plano once again asked the court to exclude the other side’s expert testimony under Daubert. Nei- ther party contends that the judge performed these analyses incorrectly. 4 More precisely, Daubert dealt with the standard for admitting expert scientific testimony. Daubert, 509 U.S. at 582. The Daubert standard was later extended to apply to all expert testimony in Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 141 (1999). 12 Nos. 14-1171 & 14-1189 Once an expert’s testimony is admitted, it is treated no differently than lay testimony. See United States v. Christian, 673 F.3d 702, 712 (7th Cir. 2012) (citing Federal Criminal Pat- tern Jury Instructions of the Seventh Circuit 3.07); United States v. Mansoori, 304 F.3d 635, 654 (7th Cir. 2002) (endorsing a jury instruction stating that “the fact that an expert has given an opinion does not mean that it is binding upon you or that you are obligated to accept the expert’s opinion as to the facts.”). When expert testimony has been admitted under Daubert, the “soundness of the factual underpinnings of the expert’s analysis and the correctness of the expert’s conclu- sions based on that analysis are factual matters to be deter- mined by the trier of fact.” Stollings v. Ryobi Tech., Inc., 725 F.3d 753, 765 (7th Cir. 2013) (quoting Smith v. Ford Motor Co., 215 F.3d 713, 718 (7th Cir. 2000)). In a bench trial, once the court has fulfilled its gatekeeping function, it becomes a trier of fact that needs to assess the evidence itself—not just the methodology underlying that evidence. The judge, as trier of fact, was required to play his “essential role as the arbiter of the weight and credibility of expert testimony.” Id. In con- cluding that Plano’s experts were more credible than Vec- chio, the judge properly scrutinized each expert’s assump- tions, reasoning, and conclusions, as is the duty of the trier of fact. Next, appellants argue that the district court was wrong to believe that they had made an argument regarding the proper securing and lashing of the molds within the ship- ping container. Appellants claim that they never made such an argument; rather, they only argued that the weight of the molds was not sufficiently dispersed. Appellants also claim that the district court was incorrect when it stated that “Dr. Vecchio’s theory of the accident hinges on the idea that the Nos. 14-1171 & 14-1189 13 Molds were not lashed down properly in the Container.” Kawasaki II, No. 07 C 5675, slip op. at 17. If appellants are correct, it is possible that the district court’s mistake caused it to wrongly discredit Vecchio’s testimony. But appellants are wrong about their own argument. In their Post-Trial Closing Statement, appellants devoted more than a page to this argument, stating that “the post- derailment investigation did not unearth any evidence” that the molds were secured and that the “‘dents’ Vecchio found in the plywood floor, however, show the steel molds were not properly secured against vertical movement.” To sup- port their lashing argument, they then quoted from Vecchio’s testimony, demonstrating that Vecchio’s theory indeed as- sumed that the crates were not lashed. Appellants’ com- plaints about the district court, therefore, are not credible. Even more disturbing is appellants’ criticism of the district court for using the word “bouncing” to refer to Vecchio’s tes- timony about “dynamic amplification.” In their closing ar- gument, appellants thrice referred to dynamic amplification as bouncing; in fact, they did not use the correct term, “dy- namic amplification,” at all. 5 5 Appellants also take issue with the district court’s statement that it was “assum[ing], without finding, that Plano is subject to Clause 10.2.” Kawa- saki II, No. 07 C 5675, slip op. at 28. Appellants argue that this was a set- tled issue, and therefore that no assumption was necessary. Though this assumption was in the appellants’ favor, and therefore did not impact the outcome below, we think it is worthwhile to provide some clarifica- tion on this matter. The appellants contend that, by questioning whether Plano was “subject to” Clause 10(2), the district court contradicted its July 19, 2013 ruling that Plano was bound to the World Bill of Lading. We appreciate the appellants’ point, as we think that the district court’s discussion here was imprecise. Clause 10(2) has a condition precedent; it 14 Nos. 14-1171 & 14-1189 We now turn to the heart of this appeal: whether the dis- trict court clearly erred by finding that appellants failed to prove that Plano breached Clause 10(2) of the World Bill of Lading. Appellants challenge both the district court’s alloca- tion of the burden of proof and its determination that appel- lants did not meet their burden. We review de novo a district court’s allocation of the burden of proof. Chi. Prime Packers, Inc. v. Northam Food Trading Co., 408 F.3d 894, 898 (7th Cir. 2005). We review a district court’s findings of fact under the clearly erroneous standard. Fed. R. Civ. P. 52(a)(6); United States v. U.S. Gypsum Co., 333 U.S. 364, 395 (1948) (“A finding is ‘clearly erroneous’ when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.”). Normally, the party claiming breach of a warranty under a maritime contract—here, appellants—bears the burden of proof. See, e.g., Cent. Oil Co. v. M/V Lamma-Forest, 821 F.2d 48, 49 (1st Cir. 1987). Appellants argue, however, that in this only applies if the “Carrier receives the goods already packed into con- tainers.” At trial, the parties disputed whether this condition had been satisfied. The district court, though, did not decide the issue because it concluded that, even if Clause 10(2) was triggered, Plano had not breached the warranty found in that Clause. As we explain below, this conclusion was sound. At that point in the litigation, it was settled that Plano was subject to Clause 10(2) because it was bound as a party to the terms of the World Bill of Lading. Rather, the question was whether the Clause was applicable in this case— that is, whether Plano’s duty to safely and properly stow the molds was triggered. This is clearly what the dis- trict court meant to say. Regardless, neither the misstatement nor the district court’s decision not to definitively rule on the matter caused ap- pellants any prejudice. Nos. 14-1171 & 14-1189 15 case the burden should fall on Plano to prove that the molds were packed properly because Plano has readier access to knowledge about the facts in question—how the shipping container was actually packed in China. Specifically, appel- lants point to Plano’s contractual relationships with World and THI, either of which could have inquired with its agents—Haixing and Shanghai Ocean Tally—about how the containers were loaded. They also argue that Plano’s engi- neering vice-president has a home in China, and that he should have investigated how the container was actually loaded. Appellants’ argument is based on John Henry Wigmore’s rule that “the burden of proving a fact is said to be put on the party who presumably has peculiar means of knowledge enabling him to prove its falsity if it is false.” 9 Wigmore, Ev- idence § 2486, at 275 (3d ed. 1940). The Supreme Court has also endorsed this rule: “the ordinary rule, based on consid- erations of fairness, does not place the burden upon a liti- gant of establishing facts peculiarly within the knowledge of his adversary.” Campbell v. United States, 365 U.S. 85, 96 (1961). Wigmore’s “rule” is not mandatory; rather, it is a “policy consideration” to be “applied in certain cases” as “a rule of fairness.” Erving Paper Mills v. Hudson-Sharp Machine Co., 332 F.2d 674, 678 (7th Cir. 1964); see also Wigmore, supra, at 275 (“The truth is that there is not and cannot be any one general solvent for all cases. It is merely a question of policy and fairness based on experience in the different situa- tions.”). Appellants point to two cases in which this maxim was employed. In Erving Paper Mills, a products liability case, we said that the manufacturer of a product should bear the bur- 16 Nos. 14-1171 & 14-1189 den of proving that the product was suitable for sale to oth- ers in the ordinary course of its business. 332 F.2d at 678. And in Shanghai Automation Instrument Co., Ltd. v. Kuei, 194 F. Supp. 2d 995, 1004 (N.D. Cal. 2001), a district court in Cali- fornia held that the burden of proof should shift to the de- fendants where the plaintiff alleged that certain funds re- ceived by the defendants were improperly converted for personal use. The defendants, the court noted, had sole pos- session of information regarding how the funds were spent. We found a number of other cases in which this rule was employed. For example, in Shatterproof Glass Corp. v. Libbey- Owens-Ford Co., the Sixth Circuit shifted the burden of proof to the defendant, which had peculiar knowledge of the roy- alty rates it had paid to various patent holders. 482 F.2d 317, 324 (6th Cir. 1973). And in PepsiCo, Inc. v. Redmond, a trade secrets case brought by Pepsi against a former employee who had joined a rival corporation, the district court granted Pepsi’s requested injunction barring Redmond from divulg- ing trade secrets. 1996 WL 3965, at *1 (N.D. Ill. Jan. 2, 1996). The court forgave Pepsi’s inability to prove what job duties Redmond would have at his new job, shifting the burden to Redmond on this issue because the details of his job duties were peculiarly within his and his new employer’s knowledge. Id. at *21. These cases are representative of the others that we examined. For a number of reasons, we decline to apply this type of burden shifting in this case. First, the information at issue is not “peculiarly within the knowledge” of Plano. Campbell, 365 U.S. at 96. In each of the above-cited cases where the burden of proof was shifted, the adversary itself had peculi- ar knowledge of the information at issue. Here, in contrast, not only does Plano lack “peculiar” (i.e., “exclusive”) Nos. 14-1171 & 14-1189 17 knowledge of the information at issue, it lacks any knowledge whatsoever—Plano simply has no information about how the shipping container was actually loaded. Ap- pellants argue that Plano has a closer relationship with the foreign companies that possess this knowledge, and there- fore that Plano has easier access to that information. Even if that is true, those companies are separate entities from Pla- no, and the mere fact that Plano may have a closer relation- ship with them does not bring this information peculiarly within Plano’s knowledge. Moreover, appellants have not demonstrated that Plano would actually be able to acquire the information at issue. The mere fact that Plano has, in the past, had contractual relationships with the foreign compa- nies that possess the relevant information does not mean that Plano currently has unfettered access to that infor- mation. It is similarly irrelevant that a Plano vice-president has a personal home in the country where those companies are based. China is vast (over 9.5 million square kilometers) and has the largest population on earth; appellants fails to explain how Plano’s vice-president would have easy access to the information at issue. See Central Intelligence Agency, The World Factbook: China (2014), available at https://www.cia.gov/library/publications/the-world- factbook/geos/ch.html (last visited Mar. 3, 2015). In sum, fairness does not dictate that the burden of proof on this is- sue be shifted to a party with no actual knowledge of the relevant information and, it seems, no ready way to acquire it. The district court, therefore, properly held appellants to the burden of proving by a preponderance of the evidence that Plano breached the warranty in Clause 10(2). That bur- den required appellants to prove it was more likely than not 18 Nos. 14-1171 & 14-1189 that Plano breached the warranty. See Bunge Corp. v. Carlisle, 227 F.3d 934, 937 (7th Cir. 2000). The district court did not clearly err by finding that appellants failed to meet this bur- den. 6 Appellants presented no evidence of how the molds were actually packed into the shipping container. The fact that only one relatively small pallet was found in the container after the crash means little—most of the container’s contents like- ly fell out during the derailment. As the district court noted, large amounts of wood—and even some complete pallets— were found near the derailment site, any of which could have come from the container in question. Appellants argue that it was “wild speculation” for the district court to sug- gest that the wood found at the crash site may have support- ed Plano’s molds. But appellants overlook the fact that they had the burden of proving that the crates were improperly supported. Because they had no direct evidence of how the container was loaded, appellants had to resort to circumstan- tial evidence—the fact that only one small pallet was found in the container post-derailment. Appellants argued that an inference should be made from this discovery that the pallet in the container was the only one used to support the molds, or that there were at most two pallets of identical size. That inference was weakened by the fact that there was ample wood found at the crash site that could have served as sup- port for the molds. The district court did not find that these scraps had supported the molds. Rather, it found that the 6 Appellants do not appeal the district court’s finding regarding how the molds were lashed and secured within the container; rather, they claim that they never made an argument about lashing in the first place. There- fore, we not need review that aspect of the district court’s opinion. Nos. 14-1171 & 14-1189 19 existence of the wood undermined appellants’ argument that the molds had been improperly supported. Moreover, Vec- chio’s assumption that the molds were supported by two identical and insufficiently supportive pallets was based on pure speculation. The missing pallet (or pallets) could just as easily have been much larger, or, conversely, nonexistent; there was simply no way to know based on what was found in the container post-derailment. Appellants repeatedly complain that Plano presented no evidence to suggest that the molds were properly packed. As the defendant, though, Plano did not have to prove anything. Appellants have not produced enough evidence to prove that Plano breached the warranty found in Clause 10(2). Their final argument, however, is that they should not have to affirmatively prove breach, but rather that breach should be presumed based on the circumstances of this case. They argue, “[c]arriers receiving a sealed container cannot possi- bly [prove how the container was loaded]—it is a ridiculous burden to carry, hence the warranty. When ‘K’ Line proved there was no evidence anywhere of proper weight dispers- ing material for the larger crate, and without proper disper- sal the stress was excessive, its prima facie case was made.” Reply Br. 17. Appellants are wrong about the purpose of the warranty—it exists to create a contractual duty, not to shift the burden of proof. Their general contention, though, is that the nature of an accident should provide a presumption of breach because, absent the presumption, carriers will rarely be able to provide direct evidence of breach, rendering the warranty a nullity. This argument is underdeveloped, and appellants provide no case law to support this approach, but we interpret it to be an attempt to import the doctrine of res ipsa loquitur into the realm of contract law. 20 Nos. 14-1171 & 14-1189 Aside from whether the res ipsa loquitur doctrine can ev- er apply in a breach of contract case, it is clear that the doc- trine should not apply in this case. We are sympathetic to appellants’ argument that it will often be exceedingly diffi- cult for a carrier to affirmatively prove that a shipping con- tainer was improperly packed when a derailment—precisely the type of injury a warranty of proper stowage is intended to prevent—has led to the destruction of the most probative evidence. This problem, however, was known to the parties when they agreed to be bound by the World Bill of Lading. With that problem in mind, appellants could have insisted upon a provision in the Bill of Lading stating that, in a case such as this, where a derailment is caused by a merchant’s goods breaking through the floor of a shipping container, there would be a presumption that the container was mis- packed. 7 As the Bill of Lading was actually written, however, there is no such presumption, and a carrier must affirmative- ly prove breach and causation. We decline to effectively re- write the terms of the Bill of Lading by presuming breach. 7 Clause 10(5) of the World Bill states: “Merchant shall inspect containers before stuffing them and the use of the containers shall be prima facie ev- idence of their being sound and suitable for use.” Therefore, it is evident that the parties knew how to incorporate evidentiary presumptions into the Bill of Lading. The failure to do so explicitly in Clause 10(2) demon- strates that the parties did not wish to create such a presumption. The presumption included in Clause 10(5), we note, does not impact our analysis of whether Plano breached its duty under Clause 10(2). As the district court correctly stated, “[t]he question of whether the stowage and seals were safe and proper is a different question than whether the container was sound and suitable.” Kawasaki II, No. 07 C 5675, slip op. at 27–28. Nos. 14-1171 & 14-1189 21 Because we agree with the district court’s conclusion that appellants did not meet their burden of proving that Plano breached Clause 10(2), we do not need to review its findings that the shipping container was defective and that those de- fects caused the molds to fall through the container floor, ul- timately causing the derailment. Plano had no obligation to explain why the accident occurred. Once the district court found that appellants had not met their burden of proving that Plano had breached Clause 10(2), the actual cause of the accident became legally irrelevant. 8 Therefore, we agree with the district court that Plano is not liable to appellants for damages stemming from the derailment. III. Conclusion We AFFIRM the judgment of the district court. 8 The district court’s finding regarding the cause of the accident, we note, was not in itself sufficient to relieve Plano of liability. To prevail, appel- lants had to prove, in addition to breach, that Plano’s breach caused the derailment. The district court found that the defective state of the ship- ping container was a but-for cause of the accident. However, the fact that a container’s defective state was a but-for cause of a derailment does not mean that a merchant’s improper stowage of goods was not also a but-for cause. Under those circumstances, the merchant could conceivably be liable because its breach was a cause (though not the only cause) of dam- age.