Missouri Pacific Railroad Company D/B/A Union Pacific Railroad Company v. Patricia Limmer, Billye Joyce Smith and Bobbye Jean Nothnagel

Motion for Rehearing Granted in Part and Denied in Part; Reversed and Remanded; Opinions of October 5, 2004 Withdrawn and Opinion on Rehearing filed November 29, 2005 Motion for Rehearing Granted in Part and Denied in Part; Reversed and Remanded; Opinions of October 5, 2004 Withdrawn and Opinion on Rehearing filed November 29, 2005. In The Fourteenth Court of Appeals ____________ NO. 14-02-00688-CV ____________ MISSOURI PACIFIC RAILROAD COMPANY D/B/A UNION PACIFIC RAILROAD COMPANY, Appellant V. PATRICIA LIMMER, BILLYE JOYCE SMITH, AND BOBBYE JEAN NOTHNAGEL, Appellees ________________________________________________________ On Appeal from the 151st District Court Harris County, Texas Trial Court Cause No. 95-42790 ________________________________________________________ O P I N I O N   O N   R E H E A R I N G Upon consideration of the motion for rehearing filed by appellees Patricia Limmer, Billye Joyce Smith, and Bobbye Jean Nothnagel (collectively referred to herein as the ALimmers@) and the responses thereto, we conclude that the trial court did not err in ruling against appellant Missouri Pacific Railroad Company d/b/a Union Pacific Railroad Company (AUnion Pacific@) as to its preemption defense. Accordingly, we grant the Limmers= motion for rehearing to this extent, deny the remainder of the Limmers= motion for rehearing, withdraw the opinions issued in this case on October 5, 2004, and issue the following opinion in their place. In this wrongful-death action arising from a collision between decedent Billy Limmer=s truck and a train at a railroad crossing, this court must determine whether the trial court erred in rejecting Union Pacific=s defense that federal law preempts the Limmers= negligence claims. We conclude the trial court did not err in rejecting Union Pacific=s preemption defense because there is legally and factually sufficient evidence to support the trial court=s implied finding that Union Pacific failed to prove federal funds were used to install warning devices at the railroad crossing in question. We further conclude the trial court reversibly erred in submitting to the jury the Limmers= negligence claim based on Union Pacific=s alleged failure to eliminate sight restrictions along its right‑of‑way because Texas law does not recognize this claim. Accordingly, we reverse the trial court=s judgment and remand this case to the trial court for a new trial consistent with this opinion. I. Factual and Procedural Background On April 24, 1994, a train struck and killed Billy Limmer when he attempted to cross railroad tracks[1] at the Front Street grade crossing in Thorndale, Texas (AFront Street Crossing@). When this accident occurred, Georgetown Railroad Company owned this train, and Southern Pacific Railroad Company (ASouthern Pacific@) operated it. Furthermore, at the time of this accident, Union Pacific owned the railroad tracks in question. At the time of the accident, the Front Street Crossing did not have automatic warning devices, which activate when a train approaches the crossing; rather, it had passive warning devices, including what are referred to as Acrossbuck@ signsCthe familiar, white, X-shaped signs with black letters spelling out, ARAILROAD CROSSING.@ There was testimony at trial that trees and vegetation in the area of the crossing obscured visibility along the tracks at the time of the accident. Also, further down the track, there was a pile of crushed limestone to be used for construction work. Billy Limmer=s heirs, the Limmers, sued Union Pacific for negligence. The jury found that the Front Street Crossing was Aextra-hazardous@ (Question 1),[2] that Union Pacific=s negligence in failing to provide automatic signals, a flag man, Aand/or@ a stop sign was a proximate cause of the collision (Question 2),[3] and that Union Pacific=s negligence in failing to eliminate sight restrictions caused by the limestone pile Aand/or@ the vegetation was a proximate cause of the collision (Question 3).[4] The jury found that Billy Limmer=s negligence was also a proximate cause of the accident, and it assigned proportionate responsibility for the collision at eighty-five percent to Union Pacific and fifteen percent to Billy Limmer. The jury returned separate compensatory damage awards for each of the Limmers. In its judgment, the trial court applied the proportionate-responsibility percentage to the jury=s damage findings and awarded the Limmers a total amount of $8,733,458.70 plus postjudgment interest. II. Issues Presented On appeal, Union Pacific presents the following issues for review: (1) Did the trial court err in rejecting Union Pacific=s preemption defense because Union Pacific conclusively proved as a matter of law that federal funds were expended to install warning devices at the Front Street Crossing, or, alternatively, is the trial court=s implied finding that this did not occur against the great weight and preponderance of the evidence? (2) Did the trial court reversibly err by submitting Union Pacific=s alleged failure to eliminate sight restrictions as an independent basis of liability? (3) Did the trial court err by refusing to instruct the jury as to the statutory duty of care owed concerning visual obstructions at public grade crossings? (4) Did the trial court err in awarding prejudgment and postjudgment interest based on delays that were not Union Pacific=s fault? III. Standards of Review In the trial court, Union Pacific asserted the affirmative defense that certain federal regulations preempt all of the Limmers= claims in this case because federal funds were expended to install warning devices at the Front Street Crossing. Before we can determine the proper standard of review regarding this issue, we must determine whether it was tried to the bench or to the jury. The record is silent as to whether the parties agreed that this preemption issue would be tried to the bench rather than to the jury. Although the Limmers demanded a jury trial and paid the jury fee, the jury never heard any of the evidence or argument on the fact issue of whether federal funds were expended to install warning devices at the Front Street Crossing. The trial court admitted the documentary evidence regarding this issue for the court only, and the trial court heard the testimony on this issue outside the presence of the jury during the trial. The trial court did not mention this fact issue in the jury charge. The trial court stated several times that it, rather than the jury, would determine the preemption issue.[5] No party objected to this procedure or indicated to the trial court that it was standing on its right to a jury trial.[6] On appeal, the Limmers do not assert that this preemption issue had to be tried to the jury or that Union Pacific waived its preemption defense by not having its preemption proof admitted in evidence before the jury. On this record, we conclude Union Pacific and the Limmers waived any right they had to have the jury decide whether federal funds were expended to install warning devices at the Front Street Crossing. See Massey v. Galvan, 822 S.W.2d 309, 318B19 (Tex. App.CHouston [14th Dist.] 1992, writ denied) (stating that a party who has paid the jury fee and properly requested a jury waives its right to a jury trial by proceeding with a trial to the bench); Sunwest Reliance Acquisitions Group, Inc. v. Provident Nat. Assur. Co., 875 S.W.2d 385, 386B88 (Tex. App.CDallas 1993, no writ) (holding that party waived right to jury trial by allowing the trial court to conduct a bench trial without objecting or indicating to the trial court that it was standing on its perfected right to a jury trial). Although the trial court filed no findings of fact or conclusions of law regarding the preemption issue, on appeal, we presume the trial court made all findings in favor of its judgment. See Pharo v. Chambers Cty., 922 S.W.2d 945, 948 (Tex. 1996) (stating that, in a bench trial in which the trial court does not file findings of fact or conclusions of law, appellate courts presume the trial court made all findings in support of its judgment). Because a complete reporter=s record is a part of the appellate record in this case, Union Pacific may challenge both the legal and factual sufficiency of the trial court=s findings. See Roberson v. Robinson, 768 S.W.2d 280, 281 (Tex. 1989). We apply the same standards of review to these challenges as those applied in the review of jury findings. See id. In its first issue, Union Pacific attacks the legal sufficiency and factually sufficiency of the evidence to support the trial court=s implied finding that Union Pacific failed to prove federal funds were expended to install warning devices at the Front Street Crossing. This finding relates to Union Pacific=s preemption defense, an issue on which Union Pacific had the burden of proof. Therefore, for Union Pacific to succeed in its legal-sufficiency challenge, we must conclude that Union Pacific conclusively proved its preemption defense as a matter of law. See Dow Chemical Co. v. Francis, 46 S.W.3d 237, 241 (Tex. 2001). In making this determination, we must consider the evidence in the light most favorable to the challenged finding and indulge every reasonable inference that would support it. See City of Keller v. Wilson, 168 S.W.3d 802, 822 (Tex. 2005). We must credit favorable evidence if a reasonable factfinder could and disregard contrary evidence unless a reasonable factfinder could not. See id. at 827. We must determine whether the evidence at trial would enable a reasonable and fair-minded person to find the facts at issue. See id. The factfinder is the sole judge of the credibility of the witnesses and the weight to give their testimony. See id. at 819. Evidence is conclusive only if reasonable people could not differ in their conclusions. See id. at 819. As to Union Pacific=s factual-sufficiency challenge, we examine the entire record, considering both the evidence in favor of, and contrary to, the challenged finding. Cain v. Bain, 709 S.W.2d 175, 176 (Tex. 1986). After considering and weighing all the evidence, we set aside the fact finding only if it is so contrary to the overwhelming weight of the evidence as to be clearly wrong and unjust. Pool v. Ford Motor Co., 715 S.W.2d 629, 635 (Tex. 1986). The trier of fact is the sole judge of the credibility of the witnesses and the weight to be given to their testimony. GTE Mobilnet of S. Tex. v. Pascouet, 61 S.W.3d 599, 615B16 (Tex. App.CHouston [14th Dist.] 2001, pet. denied). We may not substitute our own judgment for that of the trier of fact, even if we would reach a different answer on the evidence. Maritime Overseas Corp. v. Ellis, 971 S.W.2d 402, 407 (Tex. 1998). The amount of evidence necessary to affirm a judgment is far less than that necessary to reverse a judgment. Pascouet, 61 S.W.3d at 616. Union Pacific argues this court should review the trial court=s rejection of its preemption arguments completely under a de novo standard of review because it claims preemption affects the trial court=s subject matter jurisdiction. Under the federal system, once a state establishes certain courts as courts of general jurisdictionCsuch as the district court belowCit is presumed that these courts have the power to adjudicate claims under federal statutes unless otherwise provided by state or federal law.[7] See U.S. Const. art. VI, cl. 2; Tex. Const. art. V, ' 8; Tex. Gov=t Code '' 24.007, 24.008; Howlett v. Rose, 496 U.S. 356, 367B69, 110 S. Ct. 2430, 2438B39, 110 L. Ed. 2d 332 (1990); Dubai Petroleum Co. v. Kazi, 12 S.W.3d 71, 75 (Tex. 2000). The statutes and regulations at issue in this case do not contain exclusive-jurisdiction provisions. See 23 U.S.C. ' 130 (2002); 49 U.S.C. '' 20101B20153 (1997 & Supp. 2003); 23 C.F.R. ' 646.214(b) (1999). No party has cited, and we have not found, anything in state or federal law that would deprive the trial court below of subject matter jurisdiction over the Limmers= claims in this case. Ordinarily, preemption operates as an affirmative defense to a claimant=s state law claims but does not deprive state courts of jurisdiction over those claims. Mills v. Warner Lambert Co., 157 S.W.3d 424, 425 (Tex. 2005). After reviewing the statutes and regulations at issue, we conclude that there is nothing to rebut the presumption of concurrent jurisdiction and that Union Pacific=s preemption argument is not based on one of the few statutes that require claims to be resolved exclusively in a federal forum. See Mills, 157 S.W.3d at 425B28 (holding that Federal Food, Drug, and Cosmetic Act did not deprive state courts of jurisdiction, even if it were to provide defendants with an affirmative defense barring plaintiffs= claims, because that statute did not contain an exclusive-jurisdiction provision as does the Employee Retirement Income Security Act and there was no indication of a Congressional intent contrary to the presumption of concurrent jurisdiction as with the National Labor Relations Act); Smallwood v. Ill. Cent. R. Co., 385 F.3d 568, 575B76 (5th Cir. 2004) (en banc) (noting that, in case involving state law negligence suit based on accident at railroad crossing, defendant railroad company could not remove case under federal-question jurisdiction because defendant=s preemption argument was only a defense and indicating that this defense would bar claims rather than deprive state court of jurisdiction). IV. Analysis A. Is the evidence legally and factually sufficient to support the trial court=s implied finding that Union Pacific failed to prove federal funds were expended to install warning devices at the Front Street Crossing under the 1977 Program? In its first issue, Union Pacific asserts the evidence is legally and factually insufficient to support the trial court=s implied finding that federal regulations concerning warning devices at public grade crossings do not preempt the Limmers= state law claims. See Norfolk S. Ry. Co. v. Shanklin, 529 U.S. 344, 347B558, 120 S. Ct. 1467, 1470B76, 146 L. Ed. 2d 374 (2000). As to this preemption issue, Union Pacific relies on two different programs: (1) a 1977 program that sought to make sure that all public grade crossings in Texas complied with the minimum standard of the Manual on Uniform Traffic Control Devices (referred to herein as the A1977 Program@), and (2) a 1989 program to add retroreflectorized tape to the back of crossbuck signs and to their support posts (referred to herein as the A1989 Program@). Under the Supremacy Clause of the United States Constitution, federal law preempts state law when it conflicts with or frustrates federal law. U.S. Const. art. VI, cl. 2; CSX Transp., Inc. v. Easterwood, 507 U.S. 658, 663, 113 S. Ct. 1732, 1737, 123 L. Ed. 2d 387 (1993). However, when a subject is traditionally governed by state law, a court interpreting a federal statute covering the same subject should be reluctant to find preemption unless it is Athe clear and manifest purpose of Congress.@ Id., 507 U. S. at 663B64, 113 S. Ct. at 1737 (quotations omitted). If the federal statute at issue contains an express preemption clause, then the analysis must focus on the plain meaning of the clause. Id. In 1970, Congress passed the Federal Railroad Safety Act for the purpose of promoting Asafety in every area of railroad operations and [to] reduce railroad-related accidents.@ 49 U.S.C. ' 20101 (1997).[8] This statute authorizes the Secretary of Transportation to implement regulations regarding railroad safety. See id. ' 20103(a) (Supp. 2003). It also contains an express preemption clause: Laws, regulations, and orders related to railroad safety . . . shall be nationally uniform to the extent practicable. A State may adopt or continue in force a law, regulation, or order related to railroad safety . . . until the Secretary of Transportation . . . prescribes a regulation or issues an order covering the subject matter of the State requirement. A State may adopt or continue in force an additional or more stringent law, regulation, or order related to railroad safety . . . when the law, regulation or orderC (1) is necessary to eliminate or reduce an essentially local safety . . . hazard; (2) is not incompatible with a law, regulation or order of the United States Government; and (3) does not unreasonably burden interstate commerce. Id. ' 20106 (1997 & Supp. 2003). In 1973, Congress passed the Highway Safety Act, which created the Federal Highway Crossings Program and made funds available to states for the Aconstruction of projects for the elimination of hazards of railway-highway crossings.@ 23 U.S.C. ' 130(a) (2002). To participate in the program, states were required to Aconduct and systematically maintain a survey of all highways to identify those railroad crossings which may require separation, relocation, or protective devices, and establish and implement a schedule of projects for this purpose.@ Id. ' 130(d). The projects had to include, at a minimum, signs for all railway crossings. Id. The Secretary of Transportation, through the Federal Highway Administration, promulgated various regulations implementing the crossings program. The rules found at 23 C.F.R. ' 646.214(b) (1999) specifically address the design of grade crossing improvements, and subsections (3) and (4) of that provision address the adequacy of warning devices installed under the crossings program. In Easterwood, the United States Supreme Court held that, when sections 646.214(b)(3) and (4) apply, federal law preempts state tort law. See Easterwood, 507 U. S. at 670B71, 113 S. Ct. at 1741. The Court reasoned that these regulations displace state and private decision-making authority by specifying the devices to be installed and the means by which railroads are to participate in their selection. Id. In Shanklin, the Court held that the applicability of these regulations to any given railroad crossing does not depend on Aan individualized determination of adequacy@ by a diagnostic team or federal officials; rather, the regulations apply to all crossings at which warning devices are installed using federal funds. See Shanklin, 529 U.S. at 357, 120 S. Ct. at 1476. The Court explained that it is the displacement of State law by the use of federal funds to install warning devices that has preemptive effect, not whether the installation of devices at a particular crossing meets the regulatory standards. See Shanklin, 529 U.S. at 357B58, 120 S. Ct. at 1476. Because of this standard, we imply a finding by the trial court that Union Pacific failed to prove federal funds were expended to install warning devices at the Front Street Crossing under the 1977 Program. See id; Pharo, 922 S.W.2d at 948. Union Pacific asserts there is legally and factually insufficient evidence to support this finding. The evidence before the trial court relevant to this issue consists of testimony from Wayne Heathington, Douglas Woods, and Darin Kosmak. The trial court admitted in evidence (before the court only) an affidavit of Darin Kosmak, the Railroad Liaison Manager in the Traffic Operations Division of the Texas Department of Transportation (the ADepartment@). In this affidavit, Kosmak testifies as follows: [Kosmak] . . . by virtue of [his] position at the [Department] [is] personally familiar with the matters stated [in his affidavit]. . . . In 1977 the [Department] implemented a program to improve all unsignalized public grade crossings in Texas. Between 1977 and 1981 the State of Texas received Federal Highway Funding in order to implement this program. Pursuant to Section 203 of the Federal-Aid Highway Act of 1976, all crossbuck protected crossings on the Missouri Pacific Railroad Company=s road crossing received the benefit of Federal Funds between approximately 1977 and 1981. These funds were expensed to either install or upgrade crossing protection at all crossbuck-protected crossings to consisted [sic] of two (2) reflectorized crossbucks. As evidenced by the attached Exhibit Pages Nos. 1-2, The Missouri Pacific Railroad Company agreed to participate in this program. As evidenced by Exhibit Page No. 3, the letting date for the contract to install metal crossbuck poles and signs for District 17, which include[s] [the Front Street Crossing] was December 1977. The contract price, which was a 100% Federally-funded project, for District 17 was $395,069.26. Under this program, the Secretary of Transportation determined the type of warning devices to be installed at [the Front Street Crossing]; determined [sic] the means by which The Missouri Pacific Railroad Company participated in the selection; and also allocated Federal Funds [sic] for said installation. Darin Kosmak testified before the court as follows: ! It is Kosmak=s understanding that certain documents relating to the 1977 Program that were kept by the State of Texas have been discarded, before Kosmak=s employment with the Texas Department of Transportation. Kosmak assumes this occurred in the 1980s. ! Kosmak has attached to his affidavit the documents that he does have in connection with the 1977 Program. In deposition testimony admitted before the court, Darin Kosmak testified as follows: ! Kosmak cannot not say how much money, if any, was actually spent on putting crossbucks or poles or signs or replacing crossbucks or poles or signs under the 1977 Program at the Front Street Crossing. Kosmak=s affidavit does not contain any evidence specific to the Front Street Crossing regarding the expenditure of federal or state funds on that crossing under the 1977 Program. Kosmak does not know how much federal or state money, if any, was spent at the Front Street Crossing as a result of the 1977 Program. ! It was the State of Texas that made the decision to do whatever was done at the Front Street Crossing as a result of the 1977 Program, if anything. ! Kosmak does not know whether anything actually occurred at the Front Street Crossing as a result of the 1977 Program. ! Under the 1977 Program, the State of Texas would have the work done and then the federal government would reimburse the State. Wayne Heathington, an expert retained by the Limmers, testified before the court as follows: ! As Heathington understands it, Darin Kosmak of the Texas Department of Transportation did not confirm, based on his own personal knowledge, whether federal funds had been spent to install any type of crossbuck warning device at the Front Street Crossing before Billy Limmer=s fatal collision. ! The 1977 Program was federally funded and approved by the Federal Highway Administration. ! Based on his examination of a photograph of the crossbuck signs at the Front Street Crossing, Heathington believes that the crossbuck signs either were never put in under the 1977 Program or were put in under the 1977 Program but not maintained. Douglas Woods testified before the court as follows: ! He has seen the metal reflectorized crossbuck blades that have been used under the 1977 Program. ! The crossbuck blades shown in a picture of the Front Street Crossing appear to Woods to be the same type of blades. The crossbuck blades shown in a picture of the Front Street Crossing look the same as ones that Woods has seen before. ! Woods is not aware of any Union Pacific program to put in metal reflectorized crossbuck blades. A December 21, 1976 letter from the Department to Missouri Pacific Railroad Company (hereinafter AMissouri Pacific@) states as follows: [The Department is] developing a project [the 1977 Program] in cooperation with the Federal Highway Administration utilizing Safety Funds under section 203 of the Federal-Aid Highway Act of 1976 to install passive warning devices at all public road crossings of all railway lines in the State of Texas. This project is designed to bring each crossing up to the minimum standard as specified in the Manual on Uniform Traffic Control Devices. Generally, the work performed on your right of way will be the installation or upgrading of reflectorized crossbuck and number-of-track signs and placement of pavement markings and stop lines on the pavement surface, part of which may be on your right of way. . . We propose to utilize all of the existing crossbuck signs and mountings as appropriate. Where new material is installed the State will salvage and dispose of the existing signs, without credit to your company. . . . By signing and returning one copy of this letter, you grant your company=s permission for the State or its Agent to perform the work herein described as may be necessary to provide a minimum passive warning system at public highway or road crossings on your rail system in Texas. Union Pacific asserts it conclusively proved that federal funds were expended to install warning devices at the Front Street Crossing under the 1977 Program. Union Pacific bases this assertion on the following evidence: (1) Kosmak=s affidavit and portions of Kosmak=s deposition testimony; (2) the December 21, 1976 letter, (3) Woods=s testimony before the trial court; (4) Heathington=s testimony to the bench. Kosmak was not involved in the 1977 Program, and he testified at his deposition that he does not know whether anything actually occurred at the Front Street Crossing as a result of the 1977 Program or whether any federal or state money was spent at the Front Street Crossing under the 1977 Program.[9] Although the first sentence of the 1976 letter could be construed as meaning that the 1977 Program installed passive warning devices at all public road crossings in Texas, it also could reasonably be construed to mean that the 1977 Program only made sure that all public road crossings in Texas had sufficient passive warning devices. This latter interpretation is supported by the second sentence which indicates that the goal of the 1977 Program was for each crossing to conform to the minimum standard specified in the Manual on Uniform Traffic Control Devices (hereinafter the AManual@). The letter states that, generally, reflectorized crossbucks will be installed or upgraded. The letter, however, also says that existing crossbuck signs and mountings will be used, as appropriate. The letter states that, A[w]here new material is installed the State will salvage and dispose of the existing signs,@ indicating that new material will not necessarily be installed at each crossing. Finally, by signing the letter, Missouri Pacific gave its Apermission for the State or its Agent to perform the work herein described as may be necessary.@ This sentence indicates that work may not be necessary at each crossing. Union Pacific also relies on Heathington=s testimony that (1) the front of the crossbuck signs under the 1977 Program were reflectorized, like the crossbuck signs at the Front Street Crossing; (2) the 1977 Program was federally funded; and (3) federal funds were spent on the Front Street Crossing if the reflectorized crossbuck signs at that crossing were installed under the 1977 Program. This testimony does not address the issue at handCwhether federal funds were expended under the 1977 Program to install warning devices at the Front Street Crossing. Furthermore, Heathington also testified that, based on his examination of a photograph of the crossbuck signs at the Front Street Crossing, Heathington believes that the crossbuck signs at the Front Street Crossing either were never put in under the 1977 Program or were put in under the 1977 Program but not maintained. Woods, Union Pacific=s Manager of Industries and Public Projects and a longtime Union Pacific employee, did testify that the crossbuck signs shown in a picture of the Front Street Crossing appear to Woods to be the same type of signs as those that were used under the 1977 Program.[10] But, Woods was an interested witness. Testimony from interested witnesses may establish a fact as a matter of law only if the testimony could be readily contradicted if untrue, and is clear, direct, and positive, and there are no circumstances tending to discredit or impeach it. See Lofton v. Texas Brine Corp., 777 S.W.2d 384, 386 (Tex. 1989); see also City of Keller v. Wilson, 168 S.W.3d 802, 820 (Tex. 2005) (stating that the factfinder cannot ignore undisputed testimony that is clear, positive, direct, otherwise credible, free from contradictions and inconsistencies, and could have been readily controverted). Woods did not testify that any warning devices were installed at the Front Street Crossing under the 1977 Program. The record does not reflect that Woods would be in a position to have personal knowledge as to this issue. Woods simply testified the crossbuck signs at the Front Street Crossing Aappear to be the same@ and Alook the same@ as other crossbuck signs that he has seen that were installed under the 1977 Program. Woods=s testimony on this point is not clear, direct, and positive. Furthermore, there are circumstances tending to discredit or impeach Woods=s testimony. On the same day that he gave the testimony relied on by Union Pacific, Woods testified in front of the trial court and the jury that, in all his time with Union Pacific and in all his time as a driver, having driven in at least eighteen states, he has never seen a railroad crossing that he considered extrahazardous. Woods also testified that he had never seen a crossing that needed to have lights and gates not to be extrahazardous. Woods stated that in all his travels, both working for Union Pacific and personally as a motorist, he has never on any occasion ever seen a public grade crossing that had only crossbuck signs that he believed required the installation of automatic lights and gates to make the crossing safe enough for an ordinarily prudent person to cross it safely. On the previous day, Woods testified that he had been involved in between four hundred and five hundred upgrades of warning systems at Union Pacific crossings. Woods also testified that automatic lights and gates at railroad crossings do not provide additional protection for motorists against having a train/motor vehicle collision.[11] Presuming for the sake of argument that Woods=s testimony regarding the 1977 Program was uncontroverted,[12] it still does not satisfy the standard required for interested-witness testimony to be conclusive, and reasonable people could differ in their conclusions based on it. See Lofton, 777 S.W.2d at 386B87 (holding that testimony by interested witness did not establish fact as a matter of law because of circumstances tending to impeach it); Gevinson v. Manhattan Const. Co. of Ok., 449 S.W.2d 458, 467B68 (Tex. 1969) (holding testimony from interested witness did not conclusively prove reliance because there were circumstances tending to discredit or impeach that testimony). After carefully reviewing the record under the applicable standard of review, we conclude Union Pacific did not conclusively prove that federal funds were expended to install warning devices at the Front Street Crossing under the 1977 Program.[13] See Duncan v. Kansas City S. Ry. Co., 773 So.2d 670, 679B80 (La. 2000) (holding that evidence was sufficient to support trial court=s finding that railroad company failed to prove federal funds were expended to install warning devices at the railroad crossing in question because, even though there was a program using federal funds to install warning devices in the parish where the crossing was located, a letter in evidence stated that existing crossbuck signs would remain in place and there was no evidence as to whether the crossing in question had existing crossbuck signs before the project). Recognizing the trial court=s role in this case as factfinder regarding the preemption issue and applying the factual-sufficiency standard of review, we conclude that the evidence is factually sufficient to support the trial court=s implied finding that Union Pacific failed to prove federal funds were expended to install a warning device at the Front Street Crossing under the 1977 Program. Therefore, the trial court did not err in rejecting Union Pacific=s preemption defense based on the 1977 Program. See Shanklin, 529 U.S. at 357, 120 S. Ct. at 1476. B. Did the trial court err in impliedly rejecting Union Pacific=s preemption defense as to the 1989 Program because adding retroreflectorized tape to the back of crossbuck signs and their supporting pole does not constitute the installation of a warning device? The two poles that bear the crossbuck signs at the Front Street Crossing each have a band of retroreflectorized tape affixed to the pole beneath the sign, and the back sides of the crossbuck signs themselves, which contain no words or warnings, have retroreflectorized tape attached to them (we refer collectively to this tape hereinafter as the ATape@). Union Pacific asserts it conclusively proved that federal funds were expended to install the Tape and that the Tape constitutes a warning device. In the alternative, Union Pacific asserts there is factually insufficient evidence to support the trial court=s implied findings to the contrary. The Limmers assert that legally and factually sufficient evidence supports these implied findings. The Limmers also assert that sufficient evidence supports the trial court=s implied finding that the Tape was not installed during the 1989 Program. However, for the purposes of this analysis, we presume without deciding that the Tape was installed at the Front Street Crossing under the 1989 Program and that federal funds were expended on this installation.[14] As part of its burden of proving its preemption defense, Union Pacific had to show that the Tape is a warning device. See 23 C.F.R. ' 646.214(b); Shanklin, 529 U.S. at 352B53, 120 S. Ct. at 1473B74; Easterwood, 507 U.S. at 672, 113 S. Ct. at 1741. The United States Supreme Court has stated that, in making this determination, the proper inquiry is to see if the installed item meets the definition of warning device by being either an active warning device or a passive warning device as defined below: Active Warning Devices means those traffic control devices activated by the approach or presence of a train, such as flashing light signals, automatic gates and similar devices, as well as manually operated devices and crossing watchmen, all of which display to motorists positive warning of the approach or presence of a train. Passive Warning Devices means those types of traffic control devices, including signs, markings and other devices, located at or in advance of grade crossings to indicate the presence of a crossing but which do not change aspect upon the approach or presence of a train. 23 C.F.R. ' 646.204; see Easterwood, 507 U.S. at 672B73 & n.11, 113 S. Ct. at 1741B42 & n.11 (stating that installed item must meet the definition of warning device by being either an active warning device or a passive warning device as defined in 23 C.F.R. ' 646.204 and holding that motion-detection circuitry intended to be used with automatic gate at railroad crossing did not meet this definition of warning device). No party contends that the Tape is an active warning device under the above definition, and there is no evidence showing that it is. Therefore, under United States Supreme Court precedent, the issue in this case is not whether the Tape is a Awarning device@ under the ordinary meaning of that word; rather, the issue is whether the Tape constitutes a passive warning device as defined above. See Easterwood, 507 U.S. at 672B73 & n.11, 113 S. Ct. at 1741B42 & n.11; 23 C.F.R. ' 646.204. To qualify as a passive warning device, the Tape must be a Atraffic control device[] . . . located at or in advance of [the Front Street Crossing] to indicate the presence of a crossing.@ See 23 C.F.R. ' 646.204. The definition of warning device is a matter of law that has been established under precedent from the United States Supreme Court. One might therefore conclude we should determine, as a matter of law, whether any retroreflectorized tape could be a warning device under this definition or whether retroreflectorized tape installed in the 1989 Program can be such a warning device. Based on the legal definition that we must apply and based on the information that we have in the record before us regarding the Tape and the 1989 Program, we conclude that we are not in a position to determine these broader issues. Therefore, we limit our analysis to the particular tape affixed to the crossbuck signs and supporting poles at the Front Street Crossing.[15] Under the applicable legal standard, this Tape must (1) be a traffic control device, and (2) indicate the presence of a railroad crossing. See Easterwood, 507 U.S. at 672B73 & n.11, 113 S. Ct. at 1741B42 & n.11. The trial court heard the following evidence regarding these issues from the Limmers= expert, Wayne Heathington: ! Heathington is familiar with the 1989 Program to add retroreflectorized tape to crossbuck signs to increase visibility at night.[16] ! The United States Department of Transportation publishes the Manual. Nowhere in the Manual does it list or classify or describe retroreflectorized tape as a traffic control device. ! The Manual states that it presents traffic control device standards for all streets and highways open to public travel regardless of the type or class or the governmental agency having jurisdiction. ! Nowhere in the Manual does it define retroreflectorized tape as a traffic control device. ! A traffic control device has to provide information or tell the motorist something to do. The Tape reflects light so something becomes more visible and more easily seen, but the Tape has nothing to do with telling motorists to stop, to look, to observe, to turn right or to not turn right or anything of that nature. The Tape does not serve the function of a traffic control device. ! Heathington contacted the Federal Highway Administration and asked who was in charge of the Manual committee and he was told the person was Shelley Row. He then sent a letter to Ms. Row asking if the use of retroreflectorized tape on traffic sign posts and on the back of crossbuck signs is considered a traffic control device. Ms. Row responded by writing a letter as Director of the Office of Transportation Operations for the Federal Highway Administration. In this letter, Ms. Row stated that retroreflectorized tape is not considered a traffic control device. ! In enacting House Bill 2681 in 1989 (a copy of which is contained in Plaintiffs= Exhibit 110), the Texas Legislature defined an active warning device or crossbucks differently than retroreflectorized tape. The Texas Legislature made a distinction between retroreflectorized material and an active warning device or a crossbuck.[17] ! Heathington=s opinion that retroreflectorized tape cannot be considered a traffic control device is supported by House Bill 2681, which states that the State Department of Highways and Public Transportation Ashall develop guidelines and specifications for the installation and maintenance of retroreflectorized material at all public grade crossings not protected by active warning devices.@ ! H.B. 2681 defines retroreflectorized tape separately from Awarning device.@ Retroreflectorized tape is not considered to be a warning device. H.B. 2681 defines ARetroreflectorized material@ as Amaterial that reflects light so that the paths of the reflected light rays are parallel to those of the incident rays.@ H.B. 2681 defines AActive warning device@ as Aa bell, flashing light, gate, wigwag, or other automatically activated warning device.@ H.B. 2681 defines AWarning device@ as Aan active warning device, crossbuck, or other a traffic control sign, the purpose of which is to alert motorists of a grade crossing.@ ! Heathington=s opinion is also supported by article 6370b of the Texas Revised Civil Statutes, which was admitted before the trial court and which codifies H.B. 2861. Union Pacific has not cited any evidence that it introduced as to this issue, and we have found none. 1. Is the evidence sufficient to support the trial court=s implied finding that Union Pacific failed to prove that the Tape is a traffic control device? Under the applicable legal standard, to be a warning device, the Tape must be a traffic control device. See id. The regulations at issue do not define what a traffic control device is; however, they require that all traffic control devices comply with the latest edition of the Manual. The Manual states that the purpose of Atraffic control devices@ is Ato help insure highway safety by providing for the orderly and predictable movement of all traffic . . . throughout the national highway transportation system, and to provide such guidance and warnings as are needed to insure the safe and informed operation of individual elements of the traffic stream.@ The evidence from Heathington, which we presume the trial court credited, showed that the Tape is designed to make the crossbuck signs more easily visible to motorists at night and that it is not designed to control traffic. We have found no published cases dealing with the issue of whether the addition of retroreflectorized tape constitutes the installation of a warning device.[18] Furthermore, we note the conceptual difficulty of analyzing the function of the installed itemCthe TapeCapart from the warning device to which it was affixedCthe crossbuck signs.[19] While a crossbuck sign with retroreflectorized tape would be a traffic control device, the 1989 Program did not involve the installation of crossbuck signs. The testimony of Heathington and the photographs of the Tape in the record show that, although the Tape may increase the nighttime visibility of the crossbuck signs to which it is attached, the Tape, by itself, does not provide guidance or warnings to motorists. After carefully reviewing the record under the applicable standard of review, we conclude Union Pacific did not conclusively prove that the Tape is a traffic control device. See Enriquez v. Union Pac. R. Co., No. 5:03CV174, p. 23B25 (E.D. Tex. Dec. 30, 2004) (holding by Judge Folsom that Union Pacific did not prove its entitlement to summary judgment, in part, because evidence did not conclusively prove that installation of retroreflectorized tape in 1989 Program was the installation of a traffic control device or of a warning device) (not designated for publication); Lesly v. Union Pac. R. Co., No. H-03-0772, U.S. Dist. LEXIS 23018, at *11B12 (June 25, 2004) (holding by Judge Lake that Union Pacific did not prove its entitlement to summary judgment, in part, because evidence did not conclusively prove that installation of retroreflectorized tape in 1989 Program was the installation of a warning device) (not designated for publication). Applying the factual-sufficiency standard of review, we conclude that the evidence is factually sufficient to support the trial court=s implied finding that Union Pacific failed to prove that the Tape is a traffic control device. 2. Is the evidence sufficient to support the trial court=s implied finding that Union Pacific failed to prove that the Tape indicates the presence of a railroad crossing? In holding that evidence did not show that motion-detection circuitry intended to be used with an automatic gate at a railroad crossing was a warning device, the United States Supreme Court emphasized that, to meet the applicable definition of warning device, the installed item must indicate the presence of a railroad crossing. See Easterwood, 507 U.S. at 672B73 n.11, 113 S. Ct. at 1741B42 n.11. The testimony of Heathington and the photographs of the Tape in our record show that Union Pacific did not conclusively prove that the Tape, by itself, indicates the presence of a railroad crossing. Based on this evidence, reasonable people could differ in their conclusions as to whether the Tape indicates the presence of the crossbuck signs rather than indicating the presence of a railroad crossing. After carefully reviewing the record under the applicable standard of review, we conclude that Union Pacific did not conclusively prove that the Tape indicates the presence of a railroad crossing. See Enriquez, No. 5:03CV174, p. 23B25; Lesly, No. H-03-0772, U.S. Dist. LEXIS 23018, at *11B12. Applying the factual-sufficiency standard of review, we conclude that the evidence is factually sufficient to support the trial court=s implied finding that Union Pacific failed to prove that the Tape indicates the presence of a railroad crossing. Because sufficient evidence supports the trial court=s implied findings that Union Pacific did not prove that the Tape is a traffic control device and that the Tape indicates the presence of a railroad crossing, we conclude the trial court did not err in rejecting Union Pacific=s preemption defense based on the 1989 Program.[20] See Shanklin, 529 U.S. at 357, 120 S. Ct. at 1476; Easterwood, 507 U.S. at 672B73, 113 S. Ct. at 1741B42. Having found that Union Pacific=s preemption issue lack merits as to both the 1977 Program and the 1989 Program, we overrule the first issue.[21] C. Does Union Pacific=s alleged negligence in failing to eliminate sight restrictions constitute an independent basis of liability? In its second issue, Union Pacific argues that, even if federal law does not preempt the Limmers= negligence claim in Question 3, its alleged negligence in failing to eliminate sight restrictions cannot be an independent basis of liability. See supra note 4. The Limmers assert that Union Pacific failed to preserve error as to its second issue because it did not object to the jury question regarding apportionment of responsibility. The trial court overruled Union Pacific=s objection that Question 3 is not an independent basis of liability. In an analogous situation (separate liability questions but a combined apportionment question), the Texas Supreme Court recently stated that it did not have to address whether the objecting party also had to object to the apportionment question to preserve error. See Romero v. KPH Consol., Inc., 166 S.W.3d 212, 229 (Tex. 2005). In so stating, the Romero court quoted a federal case that stated in dicta that such an issue is a Aclose and difficult question.@ Romero, 166 S.W.3d at 229 n. 55 (quoting Pan Eastern Exploration Co. v. Hufo Oils, 855 F.2d 1106, 1124 (5th Cir.1988)). The case cited by the Romero court also states in dicta that A[i]t seems likely that in the case of a potentially ambiguous general verdict all the complaining party must do to protect his rights is to object to the charge and the submission vel non of the questionable theory or theories. . . .@ Pan Eastern Exploration Co. v. Hufo Oils, 855 F.2d 1106, 1124 (5th Cir.1988); see Romero, 166 S.W.3d at 229 n.55. Union Pacific had to, and did, object to the trial court=s submission of Question 3. See In re A.V., 113 S.W.3d 355, 362B63 (Tex. 2003) (holding that party failed to preserve any error based on broad-form submission of a theory without evidentiary support where party did not object to any theory=s alleged lack of evidentiary support or to the submission of a broad-form question). However, the Texas Supreme Court and this court have held that the Casteel harm analysis applies to charge error even if the parties did not object to the form of the jury charge. See Harris County v. Smith, 96 S.W.3d 230, 232, 236 (Tex. 2002) (stating that A[a] timely objection, plainly informing the court that a specific element of damages should not be included in a broad-form question because there is no evidence to support its submission, therefore preserves the error for appellate review@ and holding that Casteel harm analysis applied to error in jury question 4, even though court stated that error was preserved as to this question only by objecting that there was no evidence to support the submission of one of its damage elements); Crown Life Ins. Co. v. Casteel, 22 S.W.3d 378, 387B88 (Tex. 2000) (holding that submitting invalid theories of liability in a single broad-form jury question is harmful error when it cannot be determined whether the jury based its verdict on one or more of the invalid theories and holding that party preserved error by obtaining ruling on its timely objection asserting the invalidity of the invalid theories); Wal-Mart Stores, Inc. v. Redding, 56 S.W.3d 141, 150, 156B57 (Tex. App.CHouston [14th Dist.] 2001, pet. denied) (holding that party preserved error and was entitled to Casteel harm analysis where party objected that there was no evidence to support the submission of one of several damage elements in a question, despite party=s failure object to the broad-form nature of the jury question); see also Redding, 56 S.W.3d at 156B57 (Wittig, J., dissenting) (noting that objecting party did not object to form of charge or argue broad form was not feasible). In Romero, the trial court properly submitted one liability theory to the jury, erroneously submitted another liability theory to the jury in a separate question, and the jury answered both liability questions in the affirmative. See Romero, 166 S.W.3d at 225B29. In Romero, the court=s charge used a single question regarding apportionment of responsibility. See id. The Romero court held that a Casteel harm analysis applied in assessing whether the submission of the erroneous theory was reversible error, although it did not address whether the complaining party was required to object to the apportionment question. See id. Based on the authorities cited above, we conclude that Union Pacific did not have to object to the apportionment question to be entitled to a Casteel harm analysis, if we determine that Union Pacific=s alleged negligence in failing to eliminate sight restrictions cannot be an independent basis of liability. See Smith, 96 S.W.3d at 232B36; Casteel, 22 S.W.3d at 387B88; Redding, 56 S.W.3d at 156B57. To resolve the merits of Union Pacific=s second issue, a review of a line of Texas Supreme Court cases from the late nineteenth century is essential. First, the Texas Supreme Court held several times that the failure of a railroad company to eliminate sight restrictions along its right-of-way does not necessarily constitute negligence, but that such a failure may serve as an independent basis of liability, if the fact finder determines it to be negligence after considering all the facts and circumstances. See Galveston, H. & S.A. Ry. Co. v. Michalke, 38 S.W. 31 (Tex. 1896) (upholding ruling by court of civil appeals that trial court did not err in instructing the jury regarding sight restrictions and stating whether railroad company=s permitting of sight restrictions constituted negligence was question for fact finder), denying writ of error as to Galveston, H. & S.A. Ry. Co. v. Michalke, 37 S.W. 480 (Tex. Civ. App. 1896) (finding no error in trial court=s instruction to jury that A[i]f you believe from the evidence that defendant permitted restrictions to be placed and remain upon its track and right-of-way, so as to obstruct the view of plaintiff in approaching the public crossing in the town of Weimar, and if you believe such acts, if any, constituted negligence on the part of defendant, and that they were the proximate cause of plaintiff=s injuries, and that plaintiff did not by his negligence contribute to his injuries, then you will find your verdict for the plaintiff@); Receivers of Houston & T.C. Ry. Co. v. Stewart, 17 S.W. 33 (Tex. 1891) (holding that there was sufficient evidence to support the trial court=s fact finding that railroad company committed negligence by placing boxcars on its side track that obstructed the view of the main track and that such a finding supports a negligence claim because whether railroad company negligently used its side track was a question of fact); Dillingham v. Parker, 16 S.W. 335, 335B36 (Tex. 1891) (holding trial court erred in instructing jury that a railroad company would necessarily be negligent if it allowed cars on its side track to obstruct the view of the main track near a crossing, because whether such action by the railroad company constitutes negligence under the facts and circumstances of the case is a question for the fact finder); Eames v. T. & N.O. Ry. Co., 63 Tex. 660, 663B65 (Tex. 1885) (holding plaintiff alleged facts that a jury could find to constitute negligence, in case where plaintiff alleged railroad company negligently allowed bushes to grow on its right-of-way to the extent that they could conceal cattle, despite the railroad company=s alleged knowledge of the presence of free-roaming cattle near the track that might be obscured from the train operator=s sight by the bushes).[22] The Texas Supreme Court again visited this topic in the Rogers case, in which it held the trial court erred in instructing the jury that any failure of the railroad company to eliminate sight restrictions would necessarily constitute negligence. See Missouri, K. & T. Ry. Co. of Texas v. Rogers, 40 S.W. 956, 957B58 (Tex. 1897). The Rogers court stated the trial court erroneously charged the jury that the railroad company=s failure to eliminate sight restrictions alone would necessarily constitute negligence, regardless of the care with which the railroad company operated its train. Rogers, 40 S.W. at 957. Although the Rogers court recognized its Michalke line of cases, it also cited Cordell, a New York decision holding that the existence of sight restrictions along a railroad company=s right-of-way cannot be an independent basis of liability, although it may be a circumstance material to the issues of the plaintiff=s alleged contributory negligence and the railroad company=s alleged negligence in operating the train. See id. at 958 (citing and quoting Cordell v. N.Y. Cent. & Hudson River Ry. Co., 70 N.Y. 119 (N.Y. 1877)). Though the Cordell case is contrary to the Michalke line of cases, the Rogers court quoted from Cordell=s holding and stated that the quotation properly states Texas law, subject to a qualification relating to contributory negligence that is not relevant to the issues in the case before us. See id. at 958. The Rogers court stated that the existence of sight restrictions would not give rise to a negligence claim if the train operator exercised such care in the operation of its train as a prudent person under similar circumstances, having due regard for the safety of those traveling on the highway over the railroad, would have exercised. See id. at 958. While, on the whole, the Rogers court seems to conclude that sight restrictions are not an independent basis of liability, there is some language in the opinion that might cloud the issue. See id. at 957B58. After stating that a railroad company=s failure to eliminate sight restrictions does not necessarily constitute negligence, the Rogers court notes that it is a question of fact for the jury whether, under the circumstances, the restriction constitutes negligence and whether, under the conditions existing at the time, the railroad company exercised due care in the operation of its train. See id. While this statement may indicate that sight restrictions are to be considered in determining the train operator=s negligence, if any, in operating the train, the statement may indicate to some that sight restrictions can constitute an independent basis of liability.[23] One year later in the Knight case, the Texas Supreme Court dispelled any confusion as to the meaning of its opinion in Rogers. See Int=l & G. N. Ry. Co. v. Knight, 45 S.W. 556, 557B58 (Tex. 1898). In Knight, the trial court, in its general-verdict instructions, instructed the jury it could find the railroad company liable based solely on its determination that the company=s failure to eliminate sight restrictions (1) was an act that an ordinary prudent person would not have committed under like circumstances; (2) was negligence; and (3) was the proximate cause of the decedent=s death, without any contributory negligence by the decedent. Knight, 45 S.W. at 557. Noting that Rogers had held that sight restrictions cannot constitute negligence but are merely a matter to be considered by the fact finder as to whether the operator of the train was negligent, the court stated that the railroad company=s failure to eliminate sight restrictions in Knight could not be deemed to be negligence either in law or in fact. Id. The Knight court held that, while the evidence was sufficient to support a jury finding that the railroad company=s negligence in operating its train proximately caused the decedent=s death without any contributory negligence of the decedent, a new trial was required because the jury may have found the company liable based solely on its failure to eliminate sight restrictions under the trial court=s erroneous jury instructions. Id. at 556B57. Further, the court stated that, under its Rogers precedent, the presence of sight restrictions was material only to the extent it might affect the determination of the company=s negligence in operating its train at the crossing. Id. at 557B58. Thus, although the Michalke line of cases indicated that failure to eliminate sight restrictions could constitute an independent basis of liability, in Knight, the court held that sight restrictions could not constitute an independent basis of liability, relying upon Rogers. Knight, 45 S.W. at 557B58; Rogers, 40 S.W. at 957B58. The Texas Supreme Court has not revisited this issue in the 107 years since it decided Knight; however, lower courts applying Knight have held that plaintiffs should take nothing if their only viable basis for negligence liability against a railroad company is an alleged failure to eliminate sight restrictions, because such a failure is not an independent basis of liability. See Atchison, Topeka & Santa Fe Ry. Co. v. Rubrecht, 445 S.W.2d 784, 789 (Tex. Civ. App.CFort Worth 1969, writ ref=d n.r.e.) (reversing and rendering a take-nothing judgment because, as a matter of law, speed of train could not have proximately caused plaintiff=s injuries and because failure of railroad company to eliminate sight restrictions, though found by the jury to be negligent, cannot constitute an independent basis of liability under Rogers and Knight); Wichita Falls & S. R.R. Co. v. Anderson, 144 S.W.2d 441, 443B45 (Tex. Civ. App.CEastland 1940, writ dism=d judgm=t cor.) (holding plaintiff could not recover as a matter of law because, among other things, the failure to eliminate sight restrictions is not a ground of actionable negligence under Rogers and Knight); Texas & P. Ry. Co. v. Boyle, 29 S.W.2d 927, 930 (Tex. Civ. App.CEl Paso 1930, writ dism=d) (reversing and rendering take-nothing judgment because the only ground of alleged negligence submitted to jury was failure to eliminate sight restriction, which is not itself negligence under Rogers and Knight); Texas & N.O. Ry. Co. v. Adams, 27 S.W.2d 331, 335 (Tex. Civ. App.CBeaumont 1930, writ dism=d) (reversing and rendering a take-nothing judgment because plaintiff=s other alleged grounds of negligence failed as a matter of law and because failure to eliminate sight restrictions is not an independent basis of liability under Rogers and Knight); Galveston, H.& S.A. Ry. Co. v. McCrorey, 23 S.W.2d 691, 692B95 (Tex. Comm=n App. 1930, judgm=t adopted) (reversing and rendering take nothing judgment against plaintiff because jury found no negligence in the operation of the train but only in the railroad company=s failure to eliminate sight restrictions, which the court concludes is not an independent basis of liability under Rogers and Knight). These courts are correct that under Rogers and Knight, the Texas Supreme Court has declared that the failure to eliminate sight restrictions is not an independent basis of liability. As an intermediate court of appeals, this court is bound to follow established precedent from the Texas Supreme Court. Consideration of any changes to common-law rules must be left to that higher authority. Lubbock County, Tex. v. Trammel=s Lubbock Bail Bonds, 80 S.W.3d 580, 585 (Tex. 2002); Deutsch v. Hoover, Bax & Slovacek, L.L.P, 97 S.W.3d 179, 195 (Tex. App.CHouston [14th Dist.] 2002, no. pet.).[24] The Limmers argue that Knight and Rogers do not apply to the facts before us because Union Pacific did not operate the train, it merely owned the right-of-way. In Rogers and Knight, and those cases cited thus far regarding sight restrictions, the train operator also owned the right-of-way. As a matter of logic, if, as to sight restrictions, a right-of-way owner is not subject to liability as a right-of-way owner for accidents in which it operates the train, then there is no apparent reason that an owner should be subject to liability in its capacity as owner of the same right-of-way for accidents in which another company operates the train. Presuming that the Limmers timely sued Southern PacificCthe operator of the trainCand that the Limmers have claims that are not preempted by federal law, under Rogers and Knight, they would be free to seek a determination as to whether Southern Pacific negligently operated the train at the time of the accident made the basis of this suit, giving full consideration to the sight restrictions and other conditions at the time of the accident. The fact that different companies operate the train and own the right-of-way does not alter the rationale stated in Rogers and Knight. Research reveals no Texas case that expressly states Rogers and Knight do not apply to such circumstances, and the Limmers have not cited any authority from any jurisdiction indicating the Rogers and Knight rule does not apply to such circumstances. Indeed, a court in AlabamaCwhich has the same common-law rule as that stated in Rogers and KnightCapplied that rule to a situation in which the right-of-way owner and train operator were different companies. See Nat=l Ry. Passenger Corp. v. H & P, Inc., 949 F. Supp. 1556, 1564B65 (M.D. Ala. 1996) (applying rule of law from Alabama Great S. Ry. Co. v. Johnston, 199 So.2d 840, 843B44 (Ala. 1967), that there is no independent negligence claim for failure to eliminate sight restrictions along the right-of-way, to fact pattern in which right-of-way owner and train operator were different companies). In sum, this attempt to distinguish Rogers and Knight is unpersuasive. The Limmers also attempt to distinguish Rogers and Knight by arguing they apply only to Aeconomically useful obstructions.@ This argument is also unpersuasive. In its analysis, the Rogers court cited two vegetation cases from Illinois and the Cordell case, which involved a pile of stumps and roots. See Rogers, 40 S.W. at 958 (citing Dimick, Stables, and Cordell cases); Dimick v. Chicago & Northwestern Ry. Co., 80 Ill. 338 (Ill. 1875) (considering sight restrictions caused by trees and brush on the right-of-way as a circumstance affecting determination of plaintiff=s contributory negligence); Indianapolis & St. Louis R.R. Co. v. Stables, 62 Ill. 313 (Ill. 1872) (considering sight restrictions caused by bush and shrubs in determining negligence of railroad company in operating the train); Cordell, 70 N.Y. at 121B23 (holding that sight restrictions caused by large pile of stumps and roots and other material is not an independent basis of liability but only a circumstance to be considered in determining plaintiff=s contributory negligence and the negligence of the train operator); see also LaRocco v. Penn Cent. Transp. Co., 324 N.Y.S.2d 82, 83 (N.Y. 1971) (applying Cordell to sight restriction caused by physical condition of the road and track crossing). Although Knight involved sight restrictions relating to structures built near the tracks, it also involved sight restrictions relating to piles of cotton and railroad Aties.@ See Knight, 45 S.W. at 556. Furthermore, the Texas Commission of Appeals applied Rogers and Knight to sight restrictions caused by vegetation, and Texas courts of appeals have done likewise. See McCrorey, 23 S.W.2d at 692B95 (applying Rogers and Knight to sight restrictions caused by grass and weeds along the right-of-way); Rubrecht, 445 S.W.2d at 787B89 (applying Rogers and Knight to sight restrictions caused by trees and weeds along the right-of-way); Anderson, 144 S.W.2d at 442B45 (applying Rogers and Knight to sight restrictions caused by timber and underbrush along the right-of-way); Adams, 27 S.W.2d at 335 (applying Rogers and Knight to sight restrictions caused by a hill); Oden v. Tex. & P. Ry. Co., 9 S.W.2d 367, 368B71 (Tex. Civ. App.CTexarkana 1928, no writ) (applying Rogers and Knight to sight restrictions caused by weeds and bushes along the right-of-way). The Limmers cite three Texas cases in support of their argument that Rogers and Knight apply only to Aeconomically useful obstructions.@ See St. Louis Southwestern Ry. Co. of Tex. v. Larkin, 34 S.W.2d 693, 702 (Tex. Civ. App.CDallas 1930, writ dism=d); St. Louis, S.F. & T. Ry. Co. v. Allen, 296 S.W. 950, 953 (Tex. Civ. App.CAmarillo 1927, writ dism=d); Missouri, K & T. Ry. Co. of Tex. v. King, 123 S.W. 151, 152 (Tex. Civ. App.CDallas 1909, no writ). The Allen and King cases do not support the Limmers= argument in this regard; rather, without citing Knight, the Allen and King courts state in obiter dicta that certain sight restrictions constitute an independent basis of liability. See Allen, 296 S.W. at 951B53 (stating that trial court properly submitted special issue as to whether alleged sight restriction caused by weeds and grass was independent basis of liability, without mention of whether weeds and grass were used in railroad company=s business, in case where court affirmed judgment in favor of plaintiff based on jury answers to special issues that supported several independent basis of liability, at least two of which involved the operation of the train); King, 123 S.W. at 152B53 (stating it was permissible for the trial court to charge the jury on defendant=s alleged failure to eliminate sight restrictions caused by trees and coalhouse as an independent basis of liability, without citing Knight, and in an appeal in which the court reversed and remanded for a new trial based on charge error regarding Adiscovered peril@). These obiter dicta are incorrect under Rogers and Knight. See Knight, 45 S.W. at 557B58; Rogers, 40 S.W. at 957B58. The only language that supports the Limmers= argument is more dicta found in the Larkin case. Larkin, 34 S.W.2d at 702 (stating A[o]bstructions placed by a railroad on its right of way that are necessary, either for the transaction of the business of the railroad or for its operation, and obstructions placed on the right of way by others, with permission of the railroad, to be used in connection with the business of the railroad, cannot be made an independent ground of negligence . . . .@). The Larkin court, however, did not address the propriety of the charge regarding sight restrictions, and it affirmed a judgment for the plaintiff based on special issues supporting three different alleged grounds of negligence, two of which involved the operation of the train. Id. at 696B97. Therefore, the statements in Larkin upon which the Limmers rely were not necessary to the court=s disposition of the appeal and are obiter dicta. See Edwards v. Kaye, 9 S.W.3d 310, 314 (Tex. App.CHouston [14th Dist.] 1999, pet. denied). In addition, the distinction made in Larkin as to whether the cause of the sight restriction is used in the railroad company=s business is expressly rejected in the above analysis. See McCrorey, 23 S.W.2d at 692B95; Rubrecht, 445 S.W.2d at 787B89; Anderson, 144 S.W.2d at 442B45; Adams, 27 S.W.2d at 335; Oden, 9 S.W.2d at 368B71. Moreover, other states following the rule adopted in Rogers and Knight do not make the Aeconomically useful obstruction@ distinction. See, e.g., LaRocco, 324 N.Y.S.2d at 83 (applying same rule under New York law to sight restriction caused by physical condition of the road and track crossing); Johnston, 199 So.2d at 843B44 (applying same rule under Alabama law to sight restrictions caused by vegetation); May v. S. Ry. Co., 129 S.E.2d 624, 626B28 (N.C. 1963) (applying same rule under North Carolina law to sight restrictions caused by vegetation); Cowles, 66 A. at 1023 (applying same rule under Connecticut law to sight restrictions caused by trees). There is no merit in the Limmers= attempt to distinguish Rogers and Knight based on a requirement that the sight restriction be caused by an Aeconomically useful obstruction.@ Because the alleged theory of negligence liability submitted in Question 3 is not an independent basis of liability under Texas law, we sustain Union Pacific=s second issue. The trial court=s error is reversible if it Aprobably prevented the appellant from properly presenting the case to the court of appeals.@ See Tex. R. App. P. 44.1(a)(2); Romero, 166 S.W.3d at 227. The Texas Supreme Court has stated that this type of error is reversible unless the reviewing court is reasonably certain that the jury was not significantly influenced by the issues erroneously submitted to it. See Romero, 166 S.W.3d at 227B28. The Romero court determined that the trial court=s erroneous submission of one claim that was not supported by the evidence[25] was reversible error because the jury apportioned responsibility in a single question based on its affirmative liability findings under the erroneously submitted claim and another properly submitted claim. See id. The Romero court stated that, because the jury found liability as to the improperly submitted claim, Athe jury could not conceivably have ignored that finding in apportioning responsibility.@ Id. We are not reasonably certain that the jury in this case was not significantly influenced by the liability theory that was erroneously submitted to it in Question 3; therefore, we conclude that this error requires reversal of the judgment and a new trial.[26] See id. Conclusion There is legally and factually sufficient evidence to support the trial court=s implied findings that Union Pacific failed to prove federal funds were expended to install warning devices at the Front Street Crossing under the 1977 Program and under the 1989 Program. Under Texas common law as announced by the Texas Supreme Court and followed by lower courts, Texas does not recognize the negligence claim submitted to the jury in Question 3 as an independent basis of liability. Because we are not reasonably certain that the jury in this case was not significantly influenced by the liability theory that was erroneously submitted in Question 3, we conclude that this error requires reversal of the judgment and a new trial. Accordingly, we sustain Union Pacific=s second issue, reverse the trial court=s judgment, and remand this case to the trial court for a new trial in accordance with this opinion. /s/ Eva M. Guzman Justice Judgment rendered and Opinion on Rehearing filed November 29, 2005. Panel consists of Chief Justice Hedges and Justices Frost and Guzman. [1] The tracks had previously been owned by Missouri Pacific Railroad Company, but that company was subsequently acquired by Union Pacific Railroad Company. At the time of the accident, Union Pacific owned the tracks. [2] Question 1 defined an extra-hazardous crossing as follows: A railroad grade crossing is Aextra-hazardous@ when, because of surrounding conditions, it is so dangerous that person [sic] using ordinary care cannot pass over it in safety without some warning other than the usual cross buck [sic] sign. [3] Question 2 was conditioned on an affirmative response to Question 1 and stated as follows: Was the negligence, if any, of the Missouri Pacific Railroad, through its agents or employees, a proximate cause of the collision in question? In answering this question, you may consider only the following acts of negligence, if any: (1) failure to provide automatic signals (such as flashing lights or gates); (2) failure to provide a flag man at the Front Street crossing; and/or (3) failure to install a stop sign. [4] Question 3 provided as follows: Was the negligence, if any, of the Missouri Pacific Railroad, through its agents or employees, a proximate cause of the collision in question? In answering this question, you may consider only the following acts of negligence: (1) failure to eliminate a sight restriction, if any, caused by a pile of crushed limestone at the crossing; and/or (2) failure to eliminate a sight restriction, if any, caused by vegetation at or near the crossing. [5] For example, in discussing procedural matters related to the preemption issue, the trial court stated that A[the preemption issue] is an issue that the Court is going to decide and it=s not an issue that goes to the jury . . . .@ The trial court also asked counsel, A[I]s [the resolution of the preemption issue] really that different from, let=s say, if they were arguing attorneys fees to me? That=s something that I would be considering, not the jury.@ [6] At a pretrial conference just before trial, the Limmers agreed to the trial court=s granting Union Pacific a running objection to the evidence at the jury trial, so as to not waive Union Pacific=s argument that federal law preempts all of the Limmers= claims. In the course of discussing this running objection, the trial court told Union Pacific=s counsel, AI don=t think you should have to stand up and use the word preemption at all during this case.@ In response, the Limmers= counsel stated, AThat=s fine with me, Judge.@ [7] Union Pacific cites Cadillac Insurance Co. v. L.P.C. Distributing Co., 770 S.W.2d 892, 895 (Tex. App.CSan Antonio 1989, writ denied), for the proposition that federal preemption is a challenge to the court=s subject matter jurisdiction and, thus, is always a question of law. Cadillac Insurance does not hold that all preemption arguments defeat the trial court=s jurisdiction. Rather, it deals with state law claims relating to an employee benefit plan under the Employee Retirement Income Security Act (AERISA@). Cadillac Insurance Co., 770 S.W.2d at 893. ERISA, unlike the statutes at issue in this case, contains a provision creating exclusive jurisdiction for the federal courts as to the issues involved in the Cadillac Insurance case. See 29 U.S.C. ' 1132(e) (1999). Therefore, Cadillac Insurance is not on point. See Mills v. Warner Lambert Co., 157 S.W.3d 424, 427B28 (Tex. 2005) (distinguishing ERISA cases because they relied on exclusive-jurisdiction provision in ERISA and because statute at issue contained no such provision). [8] All statutory citations are to the current version of the statute, unless otherwise noted. [9] Union Pacific asserts that, at his deposition, Kosmak testified to the actual expenditure of federal funds at the Front Street Crossing; however, the cited testimony is vague. When asked whether he could testify under oath specifically as to how much federal money, if any, was actually spent on the Front Street Crossing to upgrade or change the warning devices before Billy Limmer=s death, Kosmak stated that A[t]here were federal funds expended, but I don=t know how much specificallyCokay, how many dollars were expended.@ Based on other portions of his deposition testimony, the trial court could have determined that Kosmak did not mean that federal funds were expended on the Front Road Crossing but that federal funds were expended generally in the 1977 and/or 1989 Program and Kosmak does not know if any funds were expended on the Front Street Crossing. Even if this testimony had the meaning ascribed to it by Union Pacific, we presume the trial court credited other parts of Kosmak=s deposition testimony, in which Kosmak testifies that he does not know whether anything actually occurred at the Front Street Crossing as a result of the 1977 Program or whether any federal or state money was spent at the Front Street Crossing under the 1977 Program or the 1989 Program. See City of Keller, 168 S.W.3d at 819. [10] Union Pacific also emphasizes Woods=s testimony that he is not aware of any Union Pacific program to put in metal reflectorized crossbuck blades. However, at the time of the 1977 Program, Missouri Pacific owned the Front Street Crossing. Although Union Pacific subsequently acquired Missouri Pacific, if a railroad company installed reflectorized crossbucks or other passive warning devices at the Front Street Crossing before 1977, it would have been Missouri Pacific or one of its predecessors rather than Union Pacific. [11] In his testimony, Woods distinguished between Aadditional warning@ and Aadditional protection.@ Woods stated that automatic lights and gates provide additional warning to motorists and then insisted that this additional warning does not provide additional protection to motorists. [12] The Limmers assert that there was contrary evidence based on the 1976 letter and Heathington=s testimony. [13] Union Pacific asserts that specific, individualized information is not required for it to have conclusively proven as a matter of law that federal funds were expended to install warning devices at the Front Street Crossing under the 1977 Program. Union Pacific cites several cases in support of this proposition. None of these cases involve an appeal by the railroad company from an adverse finding by the trial court, and none of these cases address the interested-witness rule. Furthermore, the evidence in these cases was significantly different from the evidence in this case. See O=Bannon v. Union Pac. R. Co., 169 F.3d 1088, 1089B90 (8th Cir. 1999) (affirming summary judgment based on evidence that included specific evidence that a crossbuck sign and post were installed at crossing in question in federally funded project); Wagoner v. CSX Transp., Inc., 246 F.Supp.2d 1002, 1007 (N.D. Ind. 2003) (discussing summary-judgment affidavit of railroad employee at the time of the project in question who specifically testifies that he has personal knowledge that federal funds were used to install crossbuck signs at the crossing in question); Security First Bank v. Burlington No. & Santa Fe Ry. Co., 213 F.Supp.2d 1087, 1090 (D. Neb. 2002) (describing the summary-judgment evidence only in general terms, stating that the evidence shows that the crossing in question was included in a federally funded project, and stating that the evidence was similar to the evidence in the O=Bannon case); Strozyk v. Norfolk S. Corp., No. Civ.A.01-2478, 2002 WL 1226857, at *2 (E.D. Pa. June 5, 2002) (stating that summary-judgment evidence included document with a specific description of the work to be completed at the crossing in question under a federally funded program as well as affidavit testimony that crossbuck signs were installed at that crossing using federal funds), rev=d on other grounds, 358 F.3d 268 (3rd Cir. 2004); McDaniel v. Southern Pac. Transp., 932 F. Supp 163, 167B68 (N.D. Tex. 1995) (stating that there was summary-judgment evidence showing that federal funds were expended under the 1977 Program at all Southern Pacific public road crossings and not mentioning any circumstances tending to discredit or impeach the railroad company=s evidence). [14] Kosmak testified that it cost $10 to purchase retroreflectorized tape for a railroad crossing like the Front Street Crossing. [15] This approach is consistent with the approach taken by United States District Judge Sim Lake in addressing whether retroreflectorized tape allegedly installed under the 1989 Program constituted a warning device for purposes of the railroad company=s preemption defense. See Lesly v. Union Pac. R. Co., No. H-03-0772, U.S. Dist. LEXIS 23018, at *11B12 (S.D. Tex. June 25, 2004). [16] The 1989 Texas statute implementing the 1989 Program and the letter in evidence from the United States Department of Transportation regarding this program both use the term Aretroreflectorized.@ On the other hand, in some of the testimony, witnesses and lawyers use the words Areflectorized@ or Aretroreflective@ instead. For the sake of consistency, we use Aretroreflectorized@ throughout this opinion. [17] See Act of May 17, 1989, 71st Leg., R.S., ch. 269, ' 1, 1989 Tex. Gen. Laws 1212, 1213 (formerly codified at Tex. Rev. Civ. Stat. Ann art. 6370b), repealed by Act of May 1, 1995, 74th Leg., R.S., ch. 165, ' 24(a), 1995 Tex. Gen. Laws 1025, 1870. [18] In McDaniel v. Southern Pac. Transp., the district court mentions the 1989 Program, does not address whether retroreflectorized tape is a warning device, and, in the final analysis, bases its summary judgment on the 1977 Program. See 932 F. Supp. 163, 167 (N.D. Tex. 1995). [19] Other than the unpublished Enriquez and Lesly cases cited below, our research has found only one case addressing whether an addition or enhancement to a preexisting warning device is itself the installation of a warning device. See St. Louis Southwestern Ry. Co. v. Malone Freight Lines, Inc., 39 F.3d 864, 867 (8th Cir. 1994) (stating that replacement of eight-inch lenses in existing flashing lights at railroad crossing with twelve-inch lenses did not constitute the installation of a warning device). [20] We only analyze these issues as to the tape in question at the Front Street Crossing, not as to all retroreflectorized tape. It is possible that some retroreflectorized tape might meet the definition of warning device that we apply in this case. [21] Even without any deference to the statement in Shelley Row=s letter that retroreflectorized tape is not considered a traffic control device, the evidence is sufficient to support the trial court=s implied finding that the tape in question does not meet the applicable definition of warning device. Therefore, we need not and do not address the issue of the amount of deference, if any, that we should give to this statement. [22] For ease of reference, this line of cases is referred to as the AMichalke line of cases.@ [23] The Rogers court also cites the Dillingham case for the proposition that it is a fact question for the jury whether sight restrictions constitute negligence by the railroad company. 40 S.W. at 957B58. [24] Even though some states have determined that the failure to eliminate sight restrictions can be an independent basis of liability, see, e.g., Alabama Great S. Ry. Co. v. Lee, 826 So.2d 1232, 1236 (Miss. 2002) (holding that under Mississippi precedents, railroad company may be independently liable for its negligent failure to eliminate sight restrictions caused by vegetation on its right-of-way), others hold that the failure to eliminate sight restrictions does not constitute an independent basis of liability, although the sight restrictions may be considered by the fact finder in assessing the negligence of the train operator, see, e.g., Alabama Great S. Ry. Co. v. Johnston, 199 So.2d 840, 843B44 (Ala. 1967); Cowles v. New York, N.H. & H.R. Co., 66 A. 1020, 1023 (Conn. 1907); Cordell, 70 N.Y. at 119. The Texas Supreme Court has chosen the latter rule, and this court cannot change the law in this regard. See Lubbock County, Tex., 80 S.W.3d at 585; McCrorey, 23 S.W.2d at 694B95. [25] The Supreme Court has held that, in conducting a Casteel harm analysis as to charge error, appellate courts treat charge error based on an invalid liability theory the same as charge error based on a liability theory not supported by the evidence. See Romero, 166 S.W.3d at 227. [26] Because of our disposition of this case, we need not address Union Pacific=s third and fourth issues or its argument that 49 C.F.R. 213.37(b) preempts the alleged liability submitted to the jury in Question 3.