JACOBS, Circuit Judge:
John Crane, Inc. (“Crane”), a manufacturer and distributor of sealing devices that contain asbestos, appeals from a personal injury judgment in favor of Paul and Margaret Caruolo, entered following a jury trial in the United States District Court for the Southern District of New York (Sweet, /.). It is not disputed on appeal that Paul Caruolo (“Caruolo”) contracted mesothelioma following exposure to asbestos during his service as a Navy fireman aboard ships in World War II, and during his later career in a manufacturing company in Rhode Island. Caruolo died after trial.
Crane concedes that its sealing devices contain asbestos, but argues that it is entitled to judgment as a matter of law because: (i) the asbestos is encapsulated in its products and is not released into the air in harmful form by the procedures Caruolo performed; and (ii) plaintiffs failed to present any evidence showing that Caruolo was exposed to dangerous levels of asbestos fibers released from Crane’s products. In the alternative, Crane argues that it should be granted a new trial because, inter alia, the district court permitted plaintiffs’ medical expert to testify about the amount of asbestos in visible dust emanating from asbestos-containing products and about the opinions of another expert who was not called at trial. Finally, Crane argues that the district court erroneously applied Rhode Island law in determining whether Crane is jointly and severally liable. Plaintiffs cross-appeal, arguing that the district court erroneously applied New York law in computing prejudgment interest.
We conclude that: (i) there was sufficient evidence to support the jury’s finding of Crane’s liability; (ii) the district court did not abuse its discretion in denying a new trial; (iii) the district court properly applied Rhode Island law with respect to joint and several liability; but (iv) the district court erroneously applied New York law in computing prejudgment interest. Accordingly, we affirm the judgment in all respects except that we vacate and remand for re-calculation of prejudgment interest in accordance with Rhode Island law.
BACKGROUND
Plaintiffs commenced this failure-to-warn products liability suit against 25 manufacturers of asbestos-containing insulation, claiming that Caruolo’s mesothelio-ma was caused by his workplace exposure to their products. An amended complaint added Crane as a defendant. Unlike many if not all of the other defendants, Crane [50]*50does not manufacture insulation; the only asbestos-containing products that Crane manufactures are gaskets (rings used to make joints watertight) and packing (sealant material used in valves and other devices). Caruolo claimed that he was exposed to asbestos from Crane’s products when he served in the United States Navy — at the Brooklyn Navy Yard, the Norfolk Naval Shipyard, and at sea — from 1944 to 1950, and when he worked for the Rhee Elastic Corporation in Rhode Island and its successors (collectively referred to as “Rhee”), from 1950 to 1988.
Plaintiffs’ case was consolidated with asbestos-related personal-injury claims brought by four other plaintiffs in the Southern District of New York, in part because the district court found that Car-uolo’s primary exposure to asbestos occurred at various shipyards in New York. See In re Asbestos Litig., No. 93 Civ. 3752, 1998 WL 230950, at *4-*7 (S.D.N.Y. May 8, 1998). Two of the cases settled prior to trial. The remaining three cases proceeded to a jury trial beginning on September 23, 1998. Crane was a defendant in the Caruolos’ action only. During trial, but prior to verdict, the district court ruled that Rhode Island law would govern the issue of joint and several liability, and New York law would govern the issue of prejudgment interest. See Caruolo v. A C & S, Inc., No. 93 Civ. 3752, 1998 WL 730331, at *3-*4 (S.D.N.Y. Oct.16, 1998).
The jury returned a verdict in plaintiffs’ favor, assessed damages in the amount of $7,505,000, and calculated that Crane was 10% responsible for Caruolo’s injuries and the other manufacturers shared the remaining 90% of the liability. The district court subsequently reduced the jury’s damage award by the $2,079,500 in settlements that plaintiffs had received from other defendants. The resulting judgment is $5,448,242.50, including pre-judgment interest.
Prior to the entry of judgment, plaintiffs moved for reconsideration of the district court’s ruling that New York law governed the issue of prejudgment interest. The district court denied the motion, adhering to its holding that “under New York choice of law principles, the allowance of prejudgment interest is controlled by the state ‘whose law determined liability on the main claim.’ ” Caruolo v. A C & S, Inc., No. 93 Civ. 3752, 1998 WL 730331 at *4 (S.D.N.Y. Oct. 16, 1998) (quoting Entron, Inc. v. Affiliated FM Ins. Co., 749 F.2d 127, 131 (2d Cir.1984)).
After entry of judgment, Crane moved for post-trial relief, seeking: (i) judgment as a matter of law pursuant to Rule 50 of the Federal Rules of Civil Procedure or a new trial pursuant to Rule 59(a); and (ii) reconsideration of the district court’s ruling that Rhode Island law governed the issue of joint and several liability. The district court denied the motions. See Caruolo v. A C & S, Inc., No. 93 Civ. 3752, 1999 WL 147740, at *24 (S.D.N.Y. Mar.18, 1999).
First, the district court concluded that the evidence was sufficient to support liability, citing testimony by Caruolo and his shipmates that their work caused Crane’s products — containing up to 80% asbestos— to give off visible dust, and testimony by Dr. Steven Markowitz that such visible dust contains hazardous levels of asbestos. See id. at *4-*10. The court also found the evidence sufficient to support a jury finding that Crane should have known about the dangers of its asbestos products. See id. at *10 — *11.
Second, the district court concluded that no new trial was warranted, rejecting challenges based on: (i) the admission into evidence of internal studies by bankrupt asbestos manufacturers, see id. at *12; (ii) Dr. Markowitz’s expert testimony concerning asbestos fiber release, see id. at *13; (iii) the admission into evidence of a study written by a non-testifying expert, see id. at *14-*15; (iv) errors in the jury charge, see id. at *15-*16; (v) an arguably prejudicial suggestion by plaintiffs’ counsel in summation as to the amount of a damage [51]*51award, see id. at *16; and (Vi) the exces-siveness of the damage award under New York law, see id. at *16-*20.
Finally, the district court granted Crane’s motion for reconsideration of the court’s ruling that Rhode Island law governed the issue of joint and several liability, but declined to alter its decision. See id. at *20-*24. The district court held that under New York choice of law rules, Rhode Island had the most significant interest in the outcome of the action because the Caruolos are Rhode Island residents and “the situs of Caruolo’s most regular and prolonged exposure [to asbestos from Crane’s products] was Rhode Island.” Id. at *23.
This appeal followed.
DISCUSSION
A. Judgment as a Matter of Law
We review the district court’s denial of Crane’s motion for judgment as a matter of law de novo, see Norville v. Staten Island Univ. Hosp., 196 F.3d 89, 94 (2d Cir.1999), applying the same standards as the district court, see Galdieri-Ambrosini v. National Realty & Dev. Corp., 136 F.3d 276, 289 (2d Cir.1998); Stratton v. Department for the Aging, 132 F.3d 869, 878 (2d Cir.1997). Those standards are well established: “Judgment as a matter of law may not properly be granted under Rule 50 unless the evidence, viewed in the light most favorable to the opposing party, is insufficient to permit a reasonable juror to find in her favor.” Galdieri-Ambrosini 136 F.3d at 289. A court “must give deference to all credibility determinations and reasonable inferences of the jury,” and may not weigh the credibility of witnesses or otherwise consider the weight of the evidence. Id. Thus, judgment as a matter of law should be granted only if: “(1) there is such a complete absence of evidence supporting the verdict that the jury’s findings could only have been the result of sheer surmise and conjecture, or (2) there is such an overwhelming amount of evidence in favor of the movant that reasonable and fair minded [persons] could not arrive at a verdict against [it].” Id. (citations omitted) (alterations in original).
It is undisputed in this case that New York law governs Crane’s liability. Accordingly, the inquiry is whether the evidence presented at trial, when viewed in the light most favorable to the Caruolos, is insufficient under New York law to permit a reasonable juror to find that plaintiffs met their burden of proving that Crane failed to warn of the dangers of its products.
Under New York law, “[a] manufacturer has a duty to warn against latent dangers resulting from foreseeable uses of its product of which it knew or should have known.” Liriano v. Hobart Corp., 92 N.Y.2d 232, 677 N.Y.S.2d 764, 766, 700 N.E.2d 303 (1998); see also Rastelli v. Goodyear Tire & Rubber Co., 79 N.Y.2d 289, 582 N.Y.S.2d 373, 376, 591 N.E.2d 222 (1992). “[T]he inquiry in a duty to warn case ... focus[es] principally on the foreseeability of the risk and the adequacy and effectiveness of any warning.” Liriano, 677 N.Y.S.2d at 768, 700 N.E.2d 303. “Failure-to-warn liability is intensely fact-specific, including but not limited to such issues as feasibility and difficulty of issuing warnings in the circumstances; obviousness of the risk from actual use of the product; knowledge of the particular product user; and proximate cause.” Id. at 770 (citation omitted).
Crane’s arguments on appeal concern this last issue — proximate cause. Crane does not contest that Caruolo’s mesothelio-ma was caused by exposure to asbestos. Moreover, one of Crane’s witnesses, George MeKillop, admitted that at the time Caruolo worked in the Navy, Crane’s gaskets were 70-80% asbestos and Crane’s packing was 45-50% asbestos. MeKillop also testified, however, that this asbestos was encapsulated before being used in Crane’s gaskets and packing, testimony which was corroborated at trial by the [52]*521983 deposition testimony of Vance Vo-rhees, a former Crane employee. In addition, Crane’s experts testified that Crane’s products did not release harmful amounts of asbestos. Crane therefore argues that plaintiffs failed to establish proximate cause because there was no direct evidence that linked Crane’s products to the asbestos inhaled by Caruolo.
Plaintiffs presented expert testimony on the release of asbestos from various products with which Caruolo had worked, but no such testimony as to products manufactured by Crane. But as set forth below, other evidence sufficiently supported an inference that Caruolo’s work with Crane’s products produced dangerous levels of asbestos fibers. Accordingly, this Court will not disturb the jury’s verdict.
1. Caruolo’s Exposure to Dust from Crane’s Products
There was ample evidence that Caruolo’s work with Crane’s products produced visible dust, both in the Navy and at Rhee. In his videotaped deposition, which was received in evidence, Caruolo described his experiences as a fireman on three Navy ships from 1944 to 1950. Caruolo was responsible for firing the oil-fired burners in the ships’ boiler rooms, and for repairing leaks in boiler room valves and flanges. The valves directed a fluid or substance from one place to another; the flanges connected one pipe to another. Caruolo repaired leaking valves by removing and replacing the packing; he repaired leaking flanges by removing and replacing the gaskets. Caruolo testified that he used packing and gaskets manufactured by Crane, among others.
There is evidence that these repairs caused particles of the gaskets and packing to become airborne in Caruolo’s work area. Caruolo first removed the cloth or cement insulation covering the flanges and valves. Then, on flanges, Caruolo would remove the old gasket and make a new gasket by “bang[ing]” gasket sheets with a hammer. Pieces of the gasket material would become airborne during this process, eventually falling to the floor, where they would be trampled on by the crew. On valves, Caruolo would remove the packing by pulling, chipping, poking and picking it away. In the process, packing material “definitely became airborne,” often falling on Caruolo’s clothes and on the floor. Caruolo’s shipmates corroborated Caruolo’s testimony about the dust produced when removing and replacing gaskets and valve packing.
At Rhee, Caruolo started as a mix mill operator, then worked for ten years in shipping and receiving before becoming an apprentice electrician and ultimately, a licensed journeyman electrician. Caruolo’s responsibilities as an electrician included occasionally repairing boiler controls and valves. Gaskets were often around the valves, and Caruolo removed and replaced the gaskets, as he had done in the Navy. In the process, excess gasket material would land on his work bench and when the material was cleaned up, it would often become airborne and land on his clothes. The gasket material that Caruolo handled during this time was manufactured by Crane, among others.
2. Asbestos Content of Dust
Plaintiffs’ expert witnesses testified that visible dust from gaskets or packing would likely contain hazardous levels of asbestos. Dr. Albert Miller testified that although asbestos fibers are invisible to the naked eye, when a worker sees visible dust coming from asbestos-containing products, it “means that there’s a very high concentration of asbestos” in the air. Dr. Steven Markowitz then expanded on this testimony, explaining:
you really don’t get visible asbestos dust until probably five or ten fibers per cubic centimeter of air. That’s a high level. By comparison, the current allowable level by the federal government is one-tenth of one fiber per cubic centimeter of air. When you see visible asbestos dust, you are dealing with at least [53]*53five or ten fibers per cc; in other words, 50 to 100 times what is now allowed.
Dr. Markowitz farther testified that: working around visible dust creates a “high level of hazard of the asbestos-related diseases”; when a worker is exposed to a variety of asbestos products, all of which are releasing visible dust, “all those exposures contribute[] to that person’s risk” and there is no way to determine “which fiber is not responsible”; and Caruolo’s exposures to dust emanating from gaskets contributed to Caruolo’s mesothelioma, because:
[t]hey were part of the totality of the exposure. He worked with these materials ... he pulverized [them], they became airborne, and he breathed them in. So there is no way one can say they didn’t contribute. To the contrary. All of his exposures contributed to his meso-thelioma, including this one.
Dr. Markowitz conceded on cross-examination that he had never done air sampling for fiber release from asbestos products, that such sampling would be outside his field of expertise, that he had never seen gaskets or packing installed or removed, and that he was unaware of any studies associating the use of asbestos-containing gaskets and packing (as distinct from other asbestos-containing products) with the development of mesothelioma. However, on re-direct, Dr. Markowitz testified that he was familiar with two articles that discuss the results of tests involving gaskets and packing, in which Dr. James Millette measured the amount of asbestos fibers released when packing and gaskets are removed and found a range of up to one to four fibers per cubic centimeter, a level that Dr. Markowitz characterized as “clearly beyond what’s around now and is clearly hazardous.”
We think that Dr. Markowitz’s testimony about the asbestos content in visible dust emanating from asbestos-containing products, and his testimony about Dr. Mil-lette’s study, furnished an evidentiary basis for a finding that the dust described by Caruolo as emanating from Crane’s products contained hazardous levels of asbestos.
It is true that Dr. Markowitz could point to no studies measuring asbestos released from products manufactured by Crane; and on re-cross, Dr. Markowitz could not recall whether Dr. Millette’s tests involved Crane’s products, whether the tests simulated Caruolo’s use of the gaskets and packing, or whether other similar studies existed. Dr. Markowitz’s testimony was further undermined by Crane’s expert witnesses, who testified that Crane’s products did not release hazardous levels of asbestos. Dr. Michael Matteson, an aerosol technologist who studies the measurement and movement of particles in air, used a hacksaw to cut five samples of (new and aged) Crane gaskets and packing and measured a release of only .0062 asbestos fibers per cubic centimeter, which is below the level in ambient air. Dr. Matteson also pounded gasket material with a hammer, observed visible particles, but found fiber release below the level in ambient air. Finally, Dr. Matteson tested fiber release during the removal and installation of packing, again measuring asbestos levels lower than in ambient air. Dr. Victor Roggli, co-author of Pathology of Asbestos-Associated Diseases, testified that for an asbestos exposure “to be a substantial contributing factor” to mesothelioma, the product, inter alia, “needs to create levels [of asbestos] in the environment which are substantially greater than the background levels in the environment.” Based on Dr. Matteson’s test results, Dr. Roggli concluded that it was “unlikely” that Caruolo’s exposure to Crane’s gaskets and packing was a substantial contributing factor to his mesothelioma. Dr. Roggli also explained that Dr. Millette’s study was not done in time-weighted averages and that if Dr. Millette had used such averages, his test results would have been at or below the level of asbestos in the ambient air.
Dr. Markowitz’s concessions and the testimony of Crane’s experts cast doubt on [54]*54much of Dr. Markowitz’s testimony. But it was the jury’s province to assess Dr. Markowitz’s credibility in light of all of the evidence. Moreover, doubt was cast on the expert testimony offered by Crane as well. For example, on cross-examination, Dr. Matteson admitted that his tests did not replicate the pressures and temperatures applied to the materials handled by Caruolo. More significantly, Dr. Matteson took up to five minutes to make a single cut into the gasket or packing material, likely minimizing the production of dust, whereas Caruolo’s work was much less meticulous. It was up to the jury to decide which testimony to credit and discredit.
B. New Trial Motion
“A motion for a new trial ordinarily should not be granted unless the trial court is convinced that the jury has reached a seriously erroneous result or that the verdict is a miscarriage of justice.” Atkins v. New York City, 143 F.3d 100, 102 (2d Cir.1998) (quoting Lightfoot v. Union Carbide Corp., 110 F.3d 898, 911 (2d Cir.1997)). “Unlike a motion for judgment as a matter of law, a motion for a new trial may be granted even if there is substantial evidence to support the jury’s verdict.” United States v. Landau, 155 F.3d 93, 104 (2d Cir.1998). In reviewing the district court’s denial of Crane’s new trial motion, “we must view the evidence in the light most favorable to the nonmoving party,” reversing “only if the trial court’s denial ... constitutes an abuse of discretion.” Atkins, 143 F.3d at 102 (citing Gibeau v. Nellis, 18 F.3d 107, 109 (2d Cir.1994) and Dailey v. Societe Generale, 108 F.3d 451, 458 (2d Cir.1997)).
Crane argues that: (i) the district court improperly admitted speculative testimony of Dr. Markowitz regarding the amount of asbestos fibers in visible dust from asbestos products; (ii) the district court improperly permitted Dr. Markowitz to testify about Dr. Millette’s study; (iii) the district court improperly admitted “state-of-the-art” documents from other manufacturers of asbestos and asbestos insulation; (iv) the district court failed to provide requested jury instructions; and (v) plaintiffs’ summation was improper. We conclude that the district court did not abuse its discretion in denying Crane’s motion.
1. Evidentiary Issues
The first three grounds concern the admissibility of evidence. District court judges enjoy wide latitude in determining whether evidence is admissible at trial; this Court reviews evidentiary rulings under a deferential abuse of discretion standard. See Pescatore v. Pan Am. World Airways, Inc., 97 F.3d 1, 16 (2d Cir.1996).
Crane argues that Dr. Markowitz’s testimony regarding asbestos levels in visible dust was “speculative and unsupportable.” This is not a persuasive reason for granting a new trial because Dr. Miller testified that visible dust contains very high concentrations of asbestos, testimony that Crane does not challenge. So, even if Dr. Markowitz’s testimony was admitted improperly, that error cannot be said to have resulted in a “seriously erroneous result” or a “miscarriage of justice” because other, unchallenged testimony established the same facts. Moreover, Crane was permitted to cross-examine both Drs. Markowitz and Miller, and to present testimony from its own experts refuting the opinions of Drs. Markowitz and Miller.
There similarly is no ground to grant a new trial with respect to testimony about Dr. Millette’s study. Crane argues that it was error for the district court to hold the study fell within the “learned treatise” exception to the hearsay rule, see Fed.R.Evid. 803(18), because: (i) Dr. Markowitz could not have reasonably relied on the study because his conclusions differed from those of the study; (ii) the plaintiffs failed to lay a foundation establishing that the article publishing the study’s findings was itself an authoritative publication in [55]*55the relevant field of study; and (iii) Dr. Markowitz testified about the study on redirect, not direct. Crane also argues that the admission of the study was more prejudicial than probative, see Fed.R.Evid. 403, because Dr. Millette had been retained by the plaintiffs but was never called to testify. At trial, Crane never objected to Dr. Millette’s study on hearsay or prejudice grounds, thereby failing to preserve these arguments for appeal. See United States v. Inserra, 34 F.3d 83, 90 n. 1 (2d Cir.1994) (holding that where defendant objected to admissibility only on ground of authenticity, defendant failed to preserve objection based on hearsay). We therefore review this evidentiary ruling for plain error, that is, to determine if the ruling “resulted] in a miscarriage of justice” or “is an obvious instance of misapplied law.” Latsis v. Chandris, Inc., 20 F.3d 45, 49-50 (2d Cir.1994) (internal quotation marks and citation omitted); see also United States v. Hourihan, 66 F.3d 458, 463 (2d Cir.1995).
We see no plain error in the ruling that Dr. Millette’s study fell within the “learned treatise” hearsay exception. Fed.R.Evid. 803(18). Dr. Markowitz cited the study as an example of “peer reviewed literature” demonstrating “significant dust release or asbestos release from gasket material during use.” Although the study’s findings did not coincide with Dr. Markowitz’s direct testimony, his testimony about it was provoked by Crane’s asking, on cross-examination, whether Dr. Markowitz had “found any studies associating the use of asbestos-containing gaskets to the development of mesothelioma.” In light of this challenge to his direct testimony, it was reasonable for Dr. Markowitz to rely on Dr. Millette’s study to support his testimony regarding the causal connection between the release of asbestos fibers from gaskets and the development of mesothelioma. The authoritativeness of the study was adequately established through Dr. Markowitz’s testimony. See Fed.R.Evid. 803(18) (statements contained in published treatises, periodicals or pamphlets on a scientific or medical subject may be “established as a reliable authority by the testimony or admission of the witness or by other expert testimony or by judicial notice”). The fact that Dr. Mar-kowitz first mentioned the study on redirect is of no consequence.
We also conclude that the admission of Dr. Millette’s study was not overly prejudicial. Crane had ample opportunity to cross-examine Dr. Markowitz about Dr. Millette’s study, and to elicit additional testimony about the study from its own expert, Dr. Roggli. Furthermore, the district court fully complied with Fed.R.Evid. 803(18), allowing excerpts of the study to be read to the jury without distributing printed copies.
With respect to Crane’s last evidentiary challenge, it was within the district court’s discretion to admit documents relating to other manufacturers’ knowledge of the dangers of asbestos because Crane had a duty to warn about all dangers which it “should have known to exist.” Baker v. St. Agnes Hosp., 70 A.D.2d 400, 421 N.Y.S.2d 81, 85 (2d Dep’t 1979) .(internal quotation marks and citations omitted); see also George v. Celotex Corp., 914 F.2d 26, 28-29 & n. 1 (2d Cir.1990). Because these documents reflected other industry members’ knowledge of asbestos dangers — dangers which plaintiffs were permitted to argue that Crane, as an industry member, should have known to exist — the district court properly admitted the documents.
2. Jury Charge
As to Crane’s fourth ground for a new trial, Crane argues in cursory fashion that the jury charge: (i) inadequately segregated Crane from the other defendants; (ii) omitted any reference to the special characteristics of Crane’s gaskets and packing; (iii) failed to instruct the jury against imputing other defendants’ knowledge to Crane; (iv) failed to instruct that Crane had no duty to warn of the dangers of [56]*56other manufacturers’ products; (v) confused the jury as to the meaning of “substantial contributing factor” with respect to proximate cause; and (vi) erroneously charged that Crane could be liable if found reckless.
Jury instructions are erroneous if they mislead the jury or do not adequately inform the jury of the law. See Pahuta v. Massey-Ferguson, Inc., 170 F.3d 125, 135 (2d Cir.1999); Hathaway v. Coughlin, 99 F.3d 550, 552-53 (2d Cir.1996). “We review a claim of error in the district court’s jury instructions de novo and will reverse only if the appellant shows that the error was prejudicial in light of the charge as a whole.” Japan Airlines Co. v. Port Auth., 178 F.3d 103, 110 (2d Cir.1999).
None of the claims concerning the jury charge justify a new trial. The district court was not obliged to charge the jury about the specific characteristics of Crane’s products or to otherwise distinguish Crane from the other defendants; singling out a particular piece of evidence in a jury charge is highly discretionary, and the district court did not abuse that discretion by refusing to do so here. Cf. District Council 37 v. New York City Dep’t of Parks & Recreation, 113 F.3d 347, 356 (2d Cir.1997) (“[T]he district court needs to instruct the jury only as to applicable law.”); United States v. Pujana-Mena, 949 F.2d 24, 30 (2d Cir.1991) (instruction singling out character evidence “threatens to interfere with one of the quintessential functions of the jury — to determine the relative weight, if any, to be given to particular evidence”). The district court also did not err when it failed to include an instruction explicitly concerning imputation of knowledge because the court set forth the correct legal standard that Crane had a duty to warn about dangers which it “should have known” to exist. Liriano v. Hobart Corp., 92 N.Y.2d 232, 677 N.Y.S.2d 764, 766, 700 N.E.2d 303 (1998); Rastelli v. Goodyear Tire & Rubber Co., 79 N.Y.2d 289, 582 N.Y.S.2d 373, 376, 591 N.E.2d 222 (1992); Baker, 421 N.Y.S.2d at 85.
Nor was the district court required to accept Crane’s proposed instruction that Crane had no duty to warn about other manufacturers’ products; given the facts of this case, the charge was not mandatory under New York law. See, e.g., Rastelli 582 N.Y.S.2d at 376-77, 591 N.E.2d 222 (holding there is no duty to warn about another manufacturer’s product when defendant manufactured only sound products and had “no control over the production” of the defective products at issue in the case). The district court therefore did not err in rejecting Crane’s proposed charge in favor of the legally correct instruction that the duty to warn is “nondelegable” and “a defendant may not rely on others to issue an adequate warning.” See Guarnieri v. Kewanee-Ross Corp., 270 F.2d 575, 579 (2d Cir.1959); Geressy v. Digital Equip. Corp., 980 F.Supp. 640, 650 (E.D.N.Y.1997). Similarly, although the district court’s proximate cause instruction omitted the “substantial contributing factor” language that Crane requested, the charge given certainly reflected the standards for proximate cause under New York law: “It’s not necessary for a defendant’s product to be the sole or even the dominant cause of such injuries in order to be considered proximate cause.... Where more than one factor operates separately or together with others to cause an injury or disease, each may be a proximate cause if it is a substantial factor in bringing about that injury.” See Voss v. Black & Decker Mfg. Co., 59 N.Y.2d 102, 463 N.Y.S.2d 398, 403, 450 N.E.2d 204 (1983). Finally, although Crane challenges the recklessness charge (on the ground that it failed to distinguish between defendants) Crane was never alleged to have acted recklessly, and the jury was not asked to determine whether Crane acted recklessly. Rather, the jury was asked in relevant part if Crane “fail[ed] to use reasonable care in respect to the foreseeable use in preparing, mar[57]*57keting, or warning or failing to warn about its asbestos-containing products.”
3. Summation
Crane’s final ground for seeking a new trial is that plaintiffs’ summation was improper. Crane does not specify on appeal which aspects of the summation were improper, but we assume that Crane renews its argument, made in the district court, attacking plaintiffs’ suggestion of a specific damage amount. It was within the district court’s discretion to find that this suggestion did not warrant a new trial, particularly since the district court cautioned the jury that the numbers given by the Caruolos were simply a suggestion. See Lightfoot v. Union Carbide Corp., 110 F.3d 898, 912-13 (2d Cir.1997) (declining “to adopt a per se rule prohibiting counsel from suggesting a specific sum as damages.”).
C. Choice of Law: Joint and Several Liability
It is undisputed that New York law governed the issue of Crane’s liability. Crane argues that the district court erred when it failed to apply New York law, rather than Rhode Island law, to the issue of joint and several liability as well. We review the district court’s choice-of-law determinations de novo. See United States v. Funds Held in Name or for Benefit of Wetterer, 210 F.3d 96, 106 (2d Cir.2000) (citing Sheldon v. PHH Corp., 135 F.3d 848, 852 (2d Cir.1998)).
Given the parties’ domiciles, as discussed below, this issue could be governed by the law of one of four states: Delaware, Illinois, New York or Rhode Island. It is undisputed that both Rhode Island and Delaware law impose full joint and several liability on tortfeasors; Illinois law imposes joint and several liability only upon a tortfeasor that is at least 25% at fault, see 735 Ill. Comp. Stat. 5/2-1117; Hills v. Bridgeview Little League Assoc., 306 Ill.App.3d 13, 239 Ill.Dec. 85, 713 N.E.2d 616, 623 (Ill.App.Ct.1999); and New York law provides that a tortfeasor found to be less than 50% liable pays only its proportionate share of non-economic loss, see N.Y. C.P.L.R. 1601(1); Van Vlack v. Baker, 242 A.D.2d 704, 663 N.Y.S.2d 49, 50 (2d Dep’t 1997).
Our choice is governed by the New York choice-of-law rules first set forth in Neumeier v. Kuehner, 31 N.Y.2d 121, 335 N.Y.S.2d 64, 69-70, 286 N.E.2d 454 (1972). See Barkanic v. General Admin. of Civil Aviation of the People’s Republic of China, 923 F.2d 957, 963 (2d Cir.1991) (stating that New York courts “now apply the Neu-meier rules to all post-accident loss distribution rules.”). Neumeier established three rules that take into account the domicile of the parties, the conduct at issue, and the purposes of the substantive law. See Neumeier, 335 N.Y.S.2d at 70, 286 N.E.2d 454.
The first Neumeier rule governs cases between co-domiciliaries. The second rule specifies that if the parties are domiciliar-ies of different states, and the accident occurs in one of the domiciliary states, the law of that state should be applied in the absence of special circumstances. The third rule specifies that when parties are domiciliarles of different states, and the accident occurs in a non-domieiliary state, “[n]ormally, the applicable rule of decision will be that of the state where the accident occurred but not if it can be shown that displacing that normally applicable rule will advance the relevant substantive law purposes without impairing the smooth working of the multi-state system or producing great uncertainty for litigants.” Id.
We must therefore determine: (i) where the parties are domiciled; and (ii) where Caruolo’s injury occurred.
1. Parties’Domiciles
The plaintiffs were domiciled in Rhode Island throughout this action. Crane is incorporated in Delaware and its principal place of business is in Illinois. This Court [58]*58has not yet decided whether, for choice of law purposes, a corporate defendant is domiciled in its state of incorporation or in its principal place of business. We need not answer that question, however, to consider a false conflict. See Sheldon, 135 F.3d at 853 (identifying question but declining to answer it under the facts of the case because although each state triggered the application of a different Neumeier rule, the resulting choice of law was the same). Regardless of whether Crane is domiciled in Illinois or Delaware, Rhode Island law governs because, as discussed below, Caruolo’s dominant exposure to Crane’s products occurred in that state, thereby requiring application of the second Neumeier rule.
2. Location of Injury
The district court found that Rhode Island was “the situs of Caruolo’s most regular and prolonged exposure.” Caruolo, 1999 WL 147740, at *23. In support of this finding, the district court reviewed Caruolo’s testimony regarding his work history:
Caruolo was exposed to Crane’s asbestos products in the following locations: (1) on the high seas while serving aboard the U.S.S. Missouri from 1944 to 1946; (2) in the Brooklyn Navy Yard during overhauls of the U.S.S. Missouri; (3) on the high seas while serving aboard the U.S.S. Hawkins from 1947 to 1949; (4) on the high seas while aboard the U.S.S. Leary from 1949 to 1950; (5) in the Norfolk Navy Shipyard during overhauls of the U.S.S. Hawkins and the U.S.S. Leary; and (6) from 1950 to 1988 at Rhee Elastic in Rhode Island.
Id. We review the district court’s finding for clear error. See Advani Enters., Inc. v. Underwriters at Lloyds, 140 F.3d 157, 162 (2d Cir.1998).
It is unclear where Caruolo was domiciled during his Navy years. He testified that he reported for basic training in Samson, New York, so Caruolo may have been a citizen of New York at that time, in which case we can assume that he was a citizen of New York while on the high seas. However, even if Caruolo was a New York domiciliary for the five to six years of his Navy service, that period of asbestos exposure was substantially shorter than the period of his exposure to Crane’s asbestos-containing products in Rhode Island. Car-uolo testified that he did not work with Crane’s products for the first 10 years of his work at Rhee; however, he did occasionally work with Crane’s products for his remaining 28 years at Rhee. The district court’s finding that Caruolo’s most “regular and prolonged” exposure occurred in Rhode Island is not clearly erroneous.
Because we find that the district court did not err in determining the situs of Caruolo’s injury, we uphold the district court’s application of the second Neumeier rule. Regardless of Crane’s domicile, plaintiffs were domiciled in Rhode Island and the injury occurred there. It therefore was appropriate to find that Rhode Island law applies under the second Neu-meier rule.
Analysis of Rhode Island’s contacts to this case confirms that it has the greatest interest in the joint and several liability issue. Crane sold its asbestos products in Rhode Island, and plaintiffs were Rhode Island domiciliaries; so both parties have “purposefully associated themselves” with Rhode Island. Cooney v. Osgood Mach., Inc., 81 N.Y.2d 66, 595 N.Y.S.2d 919, 923, 612 N.E.2d 277 (1993). Moreover, because plaintiffs were Rhode Island domiciliaries, Rhode Island has a significant interest in ensuring that they are adequately compensated under Rhode Island’s joint and several liability regime. See Sheldon, 135 F.3d at 853-54. New York, on the other hand, has at best a lesser interest in applying its joint and several liability rule to this case because neither party is a New York domiciliary; New York’s only significant connection to this case is that some (but by no means all) of Caruolo’s asbestos exposure occurred within its borders. Under New York choice of law rules, “the [59]*59locus jurisdiction has at best a minimal interest in determining the right of recovery or the extent of the remedy.” Id. at 854 (internal quotation marks and citations omitted). Accordingly, the district court properly applied Rhode Island law to the issue of joint and several liability.
D. Choice of Law: Prejudgment Interest
Rhode Island law provides that a tort-feasor held liable for damages must pay prejudgment interest annually at the rate of 12% from the time the cause of action accrues. See R.I. Gen. Laws 1956, § 9-21-10; Merrill v. Trenn, 706 A.2d 1305, 1310-11 (R.I.1998). When calculating plaintiffs’ judgment, however, the district court applied New York law, which provides for prejudgment interest only from verdict to judgment. See N.Y. C.P.L.R. 5002; Love v. New York, 78 N.Y.2d 540, 577 N.Y.S.2d 359, 359-60, 583 N.E.2d 1296 (1991).
The district court erred in applying New York law instead of the law of Rhode Island. The district court concluded that plaintiffs’ case was controlled by an exception to the second Neumeier rule, relying on this Court’s statement in Entron, Inc. v. Affiliated FM Insurance Co., 749 F.2d 127 (2d Cir.1984), that, under New York law, the allowance of prejudgment interest is controlled by the state “whose law determined liability on the main claim.” Id. at 131. Subsequent to Entron, however, the New York Court of Appeals drew a relevant distinction between rules in conflict that are “loss-allocating” and those that are “conduct-regulating.” See Schultz v. Boy Scouts of Am., Inc., 65 N.Y.2d 189, 491 N.Y.S.2d 90, 97-98, 480 N.E.2d 679 (1985). Prejudgment interest, like other damages issues, is an aspect of loss-allocation. See id.; Cooney, 595 N.Y.S.2d at 922-24, 612 N.E.2d 277. Accordingly, the district court, which properly applied Rhode Island law to the loss-allocating issue of joint and several liability, should have applied Rhode Island law for the same reason to the loss-allocating issue of prejudgment interest.
Crane argues that this Court’s decision in Schwimmer v. Allstate Insurance Co., 176 F.3d 648 (2d Cir.1999), is to the contrary. However, although Schwimmer cites Entrón with favor and states that “[u]nder New York choice of law rules, the law of the jurisdiction that determines liability governs the award of prejudgment interest,” Schwimmer’s holding turned on: (i) the defendant’s consent to the application of New York law to the determination of liability; and (ii) the defendant’s “fail[ure] to bring to the attention of the district court the potential applicability of Florida law to the issue of prejudgment interest.” Schwimmer, 176 F.3d at 650. Given the defendant’s waiver, the court in Schwimmer had no opportunity to consider whether, under New York law, prejudgment interest should be treated like all other loss-allocating damages issues. We address that issue today and hold that New York law, as articulated in Schultz, requires that the same analysis applicable to joint and several liability, and other loss-allocating issues, also be applied to the issue of pre-judgment interest. Accordingly, the district court erred when it applied New York law instead of Rhode Island law to calculate the pre-judgment interest to be awarded to plaintiffs.
CONCLUSION
The district court’s judgment is affirmed in all respects except that we vacate the award of pre-judgment interest and remand for re-calculation in accordance with Rhode Island law.