W&T Offshore v. Fredieu

JUSTICE BLACKLOCK delivered the opinion of the Court, in which CHIEF JUSTICE HECHT, JUSTICE GREEN, JUSTICE GUZMAN, JUSTICE DEVINE, JUSTICE BUSBY, and JUSTICE BLAND joined. JUSTICE BOYD filed a dissenting opinion. JUSTICE LEHRMANN did not participate in the decision. The federal Longshore and Harbor Workers’ Compensation Act (LHWCA) provides the exclusive remedy for covered employees injured on the job. Respondent Wesley Fredieu was injured on an offshore oil rig owned by W&T Offshore. He sued W&T. Fredieu was not W&T’s employee, but the parties dispute whether he was nevertheless covered by the LHWCA as W&T’s “borrowed employee.” If he was, this suit cannot proceed, and he must pursue his claim through the LHWCA’s workers’ compensation process. Although federal courts consider borrowed-employee status a question of law, the jury in this case was asked whether Fredieu was a borrowed employee. To make that determination, the jury was instructed to consider nine factors derived from Ruiz v. Shell Oil Co. See 413 F.2d 310, 312–14 (5th Cir. 1969). The jury found that Fredieu was not W&T’s borrowed employee and awarded damages to Fredieu. The trial court determined that submission of the borrowed- employee question to the jury was improper. After making fact-findings on the Ruiz factors itself, the trial court found that the evidence supported W&T’s borrowed-employee defense. On that basis, the trial court granted W&T’s motion for judgment notwithstanding the verdict. The court of appeals reversed, rejecting the trial court’s fact-findings and holding that the borrowed- employee inquiry can be a fact question for the jury. The court of appeals reinstated the jury’s verdict in Fredieu’s favor. We disagree with some of the court of appeals’ analysis but conclude it reached the correct result. The trial court was correct that the borrowed-employee inquiry is a legal question for the court, not a fact question for the jury. In resolving that legal question on this record, however, we conclude that W&T did not carry its burden to establish Fredieu was its borrowed employee. We also affirm the jury’s award of damages for Fredieu’s future lost earning capacity. The court of appeals’ judgment is affirmed, and the case is remanded to the trial court for entry of judgment for Fredieu. Background W&T Offshore is an oil and gas producer based in Houston. Wesley Fredieu was an employee of Wood Group Production Services, Inc., an independent contractor that provides maintenance and service work on offshore drilling platforms in the Gulf of Mexico. Wood Group contracted with W&T Offshore to provide these services. Wood Group assigned Fredieu to work on a W&T platform, where he worked for over a year. In October 2011, Fredieu was sent to work on a nearby platform, where he supervised several workers who were painting and repairing handrails. While performing a safety check, Fredieu noticed a malfunctioning regulator, which regulates pressure in the gas lines on the platform. Fredieu radioed W&T’s lead operator, located on another platform, to ask how to proceed. The lead operator told Fredieu that he should remove the regulator and bring it back to him. The two began a safety analysis, which was required before potentially hazardous operations. Over the radio, the supervisor gave Fredieu instructions for removing the regulator. Before Fredieu had removed it, a pressurized pipe near the regulator broke loose, hitting his arm and knocking him to the ground. The impact fractured both bones in his left forearm, which Fredieu underwent surgery to repair. Fredieu sued W&T for negligence. In response, W&T claimed workers’ compensation benefits were Fredieu’s sole remedy because Fredieu was acting as its borrowed employee under the LHWCA. The platform where the injury occurred is on the Outer Continental Shelf in the Gulf of Mexico. Congress passed the Outer Continental Shelf Lands Act (OCSLA) in 1953, which asserted the Federal Government’s exclusive control over the Outer Continental Shelf. See 43 U.S.C. §§ 1331–1356b. The LHWCA provides workers’ compensation for certain injuries sustained by covered persons on navigable waters of the United States. See 33 U.S.C. §§ 902–03. Relevant to this case, the OCSLA applies federal law to fixed platforms on the Outer Continental Shelf and applies the LHWCA to injuries sustained by persons working on them. See 43 U.S.C. § 1333 (a)(1), (b). The LHWCA, when it applies, provides the exclusive remedy for an injured employee who seeks compensation from his employer. See 33 U.S.C. §§ 904–05. Federal courts consider a “borrowed employee” to be an employee for purposes of the LHWCA. See, e.g., West v. Kerr- McGee Corp., 765 F.2d 526, 528–30 (5th Cir. 1985). In Ruiz v. Shell Oil Co., the Fifth Circuit announced a nine-factor test for determining whether a person is a borrowed employee. 413 F.2d at 312–14. Those factors are: (1) Who has control over the employee and the work he is performing, beyond mere suggestion of details or cooperation? (2) Whose work is being performed? (3) Was there an agreement, understanding, or meeting of the minds between the original and the borrowing employer? (4) Did the employee acquiesce in the new work situation? (5) Did the original employer terminate his relationship with the employee? (6) Who furnished tools and place for performance? (7) Was the new employment over a considerable length of time? (8) Who had the right to discharge the employee? (9) Who had the obligation to pay the employee? Gaudet v. Exxon Corp., 562 F.2d 351, 355 (5th Cir. 1977) (citing Ruiz, 413 F.2d at 312–13). This multi-factor balancing test, routinely used in the Fifth Circuit, has also been employed by many state and federal courts across the country. See, e.g., Canty v. A. Bottacchi, S.A. De Navegacion, 849 F. Supp. 1552, 1557–59 (S.D. Fla. 1994); Roberts v. Northrop Grumman Ship Sys., 108 So. 3d 471, 473 (Miss. Ct. App. 2013). Not all federal courts apply these factors, however. See, e.g., White v. Bethlehem Steel Corp., 222 F.3d 146, 150 (4th Cir. 2000). The trial court here submitted the borrowed-employee question to the jury. It was submitted as a broad-form, yes-or-no question, asking only whether Fredieu was W&T’s borrowed employee at the time of the accident: At the time of the injury in question, was Wesley Fredieu the borrowed employee of W&T? Factors to consider in determining whether Mr. Fredieu was the borrowed employee of W&T include [nine Ruiz factors listed] Answer “Yes” or “No” Answer  ______________ The jury answered “No.” Although all nine Ruiz factors were listed for the jury’s consideration, the jury was not asked to make findings for any specific Ruiz factor. The jury found W&T negligent and awarded Fredieu more than $1.7 million in damages. After the verdict, W&T moved to disregard the jury’s borrowed-employee finding as immaterial, for determination of borrowed- employee status as a matter of law, for entry of a take-nothing judgment, and for judgment notwithstanding the verdict. The trial court determined that borrowed-employee status was actually a question of law to be decided by the court. It further held that, because the broad-form borrowed-employee question did not ask the jury to make findings for each Ruiz factor, Rule 279 of the Texas Rules of Civil Procedure allowed the court to make those findings itself. After making twelve written findings of fact for the Ruiz factors, the trial court held that Fredieu was W&T’s borrowed employee. The trial court granted W&T’s motions and entered judgment notwithstanding the jury’s verdict. The court of appeals reversed. 584 S.W.3d 200, 213 (Tex. App.—Houston [14th Dist.] 2018). It held that disputed fact issues related to various Ruiz factors made the question of borrowed-employee status a question of fact. It also concluded that it was proper for the jury to balance the Ruiz factors and decide the ultimate question of borrowed-employee status. On this basis, the court of appeals held that the trial court erroneously disregarded the jury’s verdict as immaterial. It further held that Rule 279 did not permit the trial court to make fact-findings related to the Ruiz factors. The court of appeals then reviewed the jury’s borrowed-employee finding for legal sufficiency, and, after considering the Ruiz factors, affirmed. It also upheld the jury’s damages award, with one justice dissenting. Discussion W&T challenges the court of appeals’ treatment of the borrowed-employee question as well as the jury’s damages award for future lost earning capacity. We address these issues in turn. The Borrowed-Employee Question of Law “When a state court hears an admiralty case, that court occupies essentially the same position occupied by a federal court sitting in diversity: the state court must apply substantive federal maritime law but follow state procedure.” Maritime Overseas Corp. v. Ellis, 971 S.W.2d 402, 406 (Tex. 1998). The parties agree that federal substantive law and state procedural law apply to this case. They disagree about how the two interact. W&T argues that the borrowed-employee inquiry is a question of law, which means juries may not answer it in either state or federal court. Fredieu and the court of appeals reason that who decides the borrowed-employee question—as between judge and jury—can sometimes be considered a procedural question, and Texas’s broad-form jury charge practice permits juries to decide similar questions. As explained below, we agree with W&T on this point. The parties agree that in federal court, “[t]he question of borrowed-employee status is a question of law for the district court to determine.” Billizon v. Conoco, Inc., 993 F.2d 104, 105 (5th Cir. 1993). They also agree that “in some cases, factual disputes must be resolved before the district court can make its legal determination.” Id. The Fifth Circuit consistently applies this approach. See, e.g., Delahoussaye v. Performance Energy Servs., L.L.C., 734 F.3d 389, 393 (5th Cir. 2013) (“Whether Andow was a borrowed employee is a question of law . . . .”). Other federal courts do the same. See, e.g., White, 222 F.3d at 150 (borrowed-employee status is “a legal standard”); In re Knudsen, 710 F. Supp. 2d 1252, 1262 (S.D. Ala. 2010) (“[T]he question of borrowed employee status is a question of law for the court to determine.”); Sobratti v. Tropical Shipping & Constr. Co., 267 F. Supp. 2d 455, 462 (D.V.I. 2003) (“Whether one is a protected borrowing employer is a question of law, to be decided by the court in the first instance.”). While the United States Supreme Court has not addressed the matter, Fredieu cites no state or federal case in which the ultimate question of borrowed-employee status was treated as a fact question for the jury. “A question which calls for a finding beyond the province of the jury, such as a question of law, may be deemed immaterial.” Spencer v. Eagle Star Ins. Co. of Am., 876 S.W.2d 154, 157 (Tex. 1994). If the jury was asked a question of law, the trial court did not err by disregarding its answer as immaterial. Fredieu acknowledges in his brief that borrowed-employee status is “generally a legal question.” He nevertheless contends that the jury was not asked a pure question of law because many of the Ruiz factors were in genuine dispute. As Fredieu sees it, the question posed to the jury submitted the Ruiz factors in a roundabout way by instructing the jury to consider them. But this ignores the question the jury was actually asked. It was asked: “At the time of the injury in question, was Wesley Fredieu the borrowed employee of W&T?” Fredieu acknowledges that this question is generally a question of law and cites no authority to the contrary. The court of appeals suggested that “[d]etermining whether borrowed employee status is treated as a question of law or fact in any given case requires analysis of evidence pertaining to individual factors.” Fredieu, 584 S.W.3d at 213. Fredieu does not vigorously defend this proposition, and the court of appeals cited no authority for it. Nor does Fredieu identify any other authority suggesting that borrowed-employee status may be a fact question depending on the evidence. Instead, all the relevant authority consistently calls borrowed-employee status a legal question. In the Fifth Circuit, the answer to this legal question is informed by the Ruiz factors, which themselves may raise fact questions for the jury. See, e.g., Brown v. Union Oil Co. of Cal., 984 F.2d 674, 677 (5th Cir. 1993). Other than the court of appeals in this case, we are pointed to no court that has held that the ultimate question is a factual one that a jury is competent to decide. Fredieu is correct that Texas procedure favors broad-form submission of jury questions “whenever feasible.” TEX. R. CIV. P. 277. Fredieu does not contend, however, that broad-form submission practice authorizes the submission of legal questions to juries. As the Fifth Circuit understands the LHWCA, submission of the Ruiz factors in a single broad-form question is not “feasible” because the balancing of the Ruiz factors and the ultimate determination of borrowed-employee status is a question of law. Submission to the jury of individual Ruiz factors may be necessary, however, if the resolution of fact issues related to one or more factors would be material to the court’s determination of the ultimate legal question. See, e.g., Brown, 984 F.2d at 677. We are provided with no convincing reason to depart from the Fifth Circuit’s approach, and Fredieu does not offer any reason to think the Fifth Circuit’s view is incorrect as a matter of statutory interpretation, although he would be free to try.[1] We agree with the Fifth Circuit that, in LHWCA cases, “the question of borrowed-employee status is a question of law for the district court to determine.” Billizon, 993 F.2d at 105. Submitting this legal question to the jury was improper, and the trial court did not err in disregarding the jury’s answer to it. On the other hand, Fredieu is correct that borrowed-employee status is a defense on which W&T bore the burden of proof. See Exxon Corp. v. Perez, 842 S.W.2d 629, 631 (Tex. 1992) (per curiam). W&T was therefore required to secure any jury findings necessary to support its defense. Phila. Indem. Ins. Co. v. White, 490 S.W.3d 468, 485 (Tex. 2016) (explaining that the burden of proof is on the defendant to present sufficient evidence to establish affirmative defenses and obtain the requisite jury findings); Franks v. Associated Air Ctr., Inc., 663 F.2d 583, 587 (5th Cir. 1981) (holding that the party asserting the borrowed-employee defense bears the burden of proof). As explained above, in the LHWCA context, “borrowed employee status is a matter of law . . . but some cases involve factual disputes on the issue of borrowed employee status and require findings by a fact-finder.” Brown, 984 F.2d at 677. W&T did not ask the trial court to submit any of the Ruiz factors to the jury. It does not contend the Ruiz factors provide the wrong framework. Instead, it agrees they must be submitted to the jury when necessary, but it admittedly never asked the trial court to do so. W&T therefore waived its entitlement to fact-findings on the Ruiz factors.[2] Having secured no fact-findings in its favor on the Ruiz factors, W&T is only entitled to have a Ruiz factor weighed in its favor if (1) there is no genuine, material dispute that the factor weighs in its favor, or (2) the evidence conclusively establishes that the factor weighs in its favor. See Phila. Indem. Ins. Co., 490 S.W.3d at 489 (explaining that, if a party fails to obtain requisite fact-findings on an affirmative defense, the evidence must conclusively establish those facts for the party to prevail on them on appeal). W&T embraces this burden to some extent, arguing that the evidence conclusively establishes the disputed Ruiz factors in its favor. As explained below, we disagree and, after analyzing all of the evidence, hold that W&T failed to prove its borrowed-employee defense.[3] B. The Ruiz Factors To analyze whether Fredieu was a borrowed employee, we look to federal authority applying the LHWCA, not state employment law. The LHWCA borrowed-employee inquiry is “best viewed as a question of the extent of coverage under the LHWCA, [so] federal law applies.” Gaudet, 562 F.2d at 357. We therefore look to LHWCA caselaw for guidance in determining when an independent contractor’s employee, like Fredieu, becomes the borrowed employee of the company to which he provides services. Although not all federal courts agree that the nine Ruiz factors provide a useful framework for determining borrowed-employee status,[4] the Fifth Circuit has consistently applied them for many years, and neither party suggests we should do otherwise. We therefore apply the nine- factor Ruiz analysis as the parties request, considering each factor before determining as a matter of law whether W&T proved Fredieu was its borrowed employee. The nine factors are: Who has control over the employee and the work he is performing, beyond mere suggestion of details or cooperation? Whose work is being performed? Was there an agreement, understanding, or meeting of the minds between the original and the borrowing employer? Did the employee acquiesce in the new work situation? Did the original employer terminate his relationship with the employee? Who furnished tools and place for performance? Was the new employment over a considerable length of time? Who had the right to discharge the employee? Who had the obligation to pay the employee? Gaudet, 562 F.2d at 355 (citing Ruiz, 413 F.2d at 312–13). “[N]o one of these factors, or any combination of them, is decisive, and no fixed test is used to determine the existence of a borrowed-servant relationship.” Ruiz, 413 F.2d at 312. Factor 1: Who had control over the employee and the work he was performing, beyond mere suggestion of the details or cooperation? This factor, while not determinative, is the “central factor” for analyzing borrowed- employee status. Brown, 984 F.2d at 676. The control factor “is perhaps the most universally accepted standard for establishing an employer-employee relationship.” Ruiz, 413 F.2d at 312. Fredieu was hired and trained by Wood Group. Wood Group assigned him to work on the W&T platform. Fredieu presented evidence showing that W&T was generally hands-off with respect to his work. He testified that W&T did not give him “constant or regular instruction,” and that W&T did not tell him how to do things or control the details of his day-to-day work. On the day of the accident, W&T personnel summarized generally what Fredieu should do by radio, but “left a bunch of steps out,” according to an expert witness. No W&T supervisor was present on the platform or dispatched to supervise Fredieu when he discovered the faulty regulator. Fredieu also points to one telling piece of trial testimony from a W&T Vice President who said that Fredieu was in control of his own work, and that W&T “trusted his judgment” in supervising other contract workers on the day of the accident. W&T argues that, while it did not control Fredieu’s day-to-day routine activities, Fredieu himself testified that he received guidance and direction from W&T for anything “that needed to be done outside of a routine activity that took place each and every day.” W&T also argues that it controlled Fredieu and the details of his work at the moment of the accident. A W&T employee was giving Fredieu instructions on how to remove the regulator when Fredieu was injured. Fredieu alerted W&T to the regulator malfunction and testified, “I have to call and let [a W&T lead operator] know what my findings were and how he wanted me to go about fixing it.” Fredieu also testified that he was doing what he was told by W&T in attempting to remove the regulator. We conclude that this conflicting evidence raises a material fact issue on the control factor, which should have been submitted to the jury. W&T did not seek such a finding; nor does the evidence conclusively establish its control of Fredieu. W&T is therefore not entitled to have the control question weighed in its favor. While W&T gave remote instructions to Fredieu at the time of the accident, this was a function of safety protocol as part of a required “Job Safety Analysis,” not the kind of day-to-day control that is crucial in considering when an employee of one entity has become the borrowed employee of another as a matter of law. Furthermore, testimony by W&T’s Vice President that W&T did not control Fredieu weighs in favor of Fredieu on this issue and at least created a material fact issue on the key question of control. Factor 2: Whose work is being performed? Fredieu agrees that he performed W&T’s work. Factor 3: Was there an agreement, understanding, or meeting of the minds between the original and the borrowing employer? Fredieu argues that there was such an agreement. He points to a portion of the master service contract (MSC) between W&T and Wood Group labeled “Independent Contractor,” which states that “[Wood Group] is an independent contractor and that neither Contractor nor Contractor’s principals, partners, employees, or subcontractors are servants, agents, or employees of W&T.” The MSC also states that an employee “shall always be the employee of the entity from which he receives his paycheck,” and Fredieu was paid by Wood Group. W&T points to other language in the MSC that required W&T to maintain workers’ compensation insurance with “appropriate Borrowed Servant or Alternate Employer endorsement(s) in favor of W&T Group and such that W&T Group shall have full coverage as to all members of Contractor Group (who may be deemed borrowed servants or statutory employees of W&T Group).” W&T also argues that “[t]he reality at the worksite and the parties’ actions in carrying out a contract . . . can impliedly modify, alter, or waive express contract provisions.” Melancon v. Amoco Prod. Co., 834 F.2d, 1238, 1245 (5th Cir. 1988). It points to the fact that Fredieu shared sleeping, showering, and eating facilities with W&T’s employees and received instructions as to which tasks he would be performing on W&T’s platforms. We are unconvinced that living arrangements on the rig impliedly modified Fredieu’s contractual relationship with W&T. W&T contractually disclaimed that Fredieu was its employee. In contexts other than the LHWCA, it would no doubt have vigorously resisted the suggestion that it had any kind of employment relationship with him. While the MSC’s requirement that W&T maintain workers’ compensation insurance for Wood Group’s employees provides considerable support for W&T’s position, it does not conclusively establish that Wood Group and W&T agreed Fredieu would be W&T’s borrowed employee. An unresolved material fact question remains on this question, and in the absence of a fact-finding in W&T’s favor, this factor cannot be weighed for W&T. Factor 4: Did the employee acquiesce in the new work situation? There is no dispute Fredieu agreed to his assignment to work on the W&T rigs. Factor 5: Did the original employer terminate his relationship with the employee? This factor asks whether the original employer had, in effect, temporarily terminated its relationship with the employee, but does not “require[] a lending employer to completely sever his relationship with the employee.” Capps v. N.L. Baroid-NL Indus., Inc., 784 F.2d 615, 617 (5th Cir. 1986). “The emphasis . . . should focus on the lending employer’s relationship with the employee while the borrowing occurs.” Id. at 618. Courts consider whether the lending employer supervises the employee’s work, has the right to relocate the employee to another location if needed, or otherwise restricts or exercises control over the employee. Mere nominal control by the lending employer indicates a termination of the relationship. See Melancon, 834 F.2d at 1246. The evidence indicated that Fredieu’s supervisor at Wood Group determined his platform assignments. In addition, both Fredieu and his supervisor testified that Wood Group could assign Fredieu to work for any platform or for other customers, and Fredieu testified that he continued to be trained by Wood Group. There is also evidence that Wood Group outfitted Fredieu in a hard hat, shirt, and jacket—all with Wood Group’s logo—and with pants, safety glasses, gloves, and boots. In response, W&T argues that this authority by Wood Group only constituted nominal control over Fredieu and points to Fredieu’s testimony that he never called his supervisor at Wood Group to receive instructions about his work for W&T. Wood Group’s ability to assign, reassign, and relocate Fredieu is some evidence favoring Fredieu on this factor. Although the evidence is mixed, its indeterminacy undermines W&T’s effort to conclusively establish that the Wood Group effectively terminated its relationship with Fredieu. Factor 6: Who furnished tools and place for performance? Fredieu concedes that W&T provided the place for performance but argues that W&T, Wood Group, and Fredieu himself all contributed tools for his work. Fredieu testified that Wood Group provided him with a hard hat, shirt, jacket, pants, safety glasses, gloves, and boots. He also stated that he provided his own crescent wrench and channel-lock pliers, which he says were the tools he used almost exclusively on jobs. W&T argues that it provided a radio and all of the specialized equipment needed for Fredieu’s work, like “tubing, fittings, hammer unions, pipe wrenches, bolts on flanges and hammer wrenches.” W&T unquestionably provided the place of performance, although most of the basic tools Fredieu regularly used were provided by Wood Group or himself. This factor favors W&T, but it makes little difference to the ultimate determination when weighed against the other factors. Factor 7: Was the new employment over a considerable length of time? Fredieu agrees that the new employment was over a considerable length of time. Factor 8: Who had the right to discharge the employee? W&T had the right to terminate Fredieu’s services. Fredieu does not challenge conclusions by the trial court and the court of appeals that this factor favors W&T. Factor 9: Who had the obligation to pay the employee? The evidence is undisputed that Fredieu’s paychecks came from Wood Group. While W&T paid Wood Group for the services provided by Wood Group’s employees, this factor favors Fredieu because, from his perspective, the employer obligated to pay him was Wood Group. There was no evidence that W&T’s obligations to pay Wood Group had to be met before Wood Group was obligated to pay Fredieu. * * * Because the ultimate question of borrowed-employee status is one of law, we approach de novo the question whether Fredieu was W&T’s borrowed employee. See, e.g., State v. Delany, 197 S.W.3d 297, 300 (Tex. 2006). We do so, however, with the obligation to weigh the Ruiz factors in light of how they were resolved—or in this case not resolved—by the fact-finder.[5] Considering all the Ruiz factors and W&T’s failure to obtain fact-findings in its favor on any of them, we conclude the evidence does not establish that Fredieu was W&T’s borrowed employee. As discussed above, some factors favor W&T, but some do not clearly favor either party, including the “central question” of control. Traditionally, control has been the key factor in determining borrowed-employee status. E.g., Standard Oil Co. v. Anderson, 212 U.S. 215, 221 (1909) (“If that other furnishes him with men to do the work and places them under his exclusive control in the performance of it, those men became pro hac vice the servants of him to whom they are furnished.”); Hebron v. Union Oil Co. of Cal., 634 F.2d 245, 247 (5th Cir. 1981) (per curiam) (“The central question in borrowed servant cases is whether someone has the power to control and direct another person in performance of his work.”); Capps, 784 F.2d at 617 (“[T]he courts place the most emphasis on the first factor, control over the employee.”). Here, W&T did not conclusively establish that it controlled Fredieu’s day-to-day work. W&T’s vice president testified that Fredieu was in control of his own work. Moreover, Wood Group maintained a substantial degree of control of Fredieu in the ability to assign him, reassign him, or relocate him to platforms or different customers altogether. Fredieu also wore a Wood Group uniform and equipment, reinforcing his separation from W&T’s employees. There is countervailing evidence indicating W&T’s control of Fredieu’s work duties. And the MSC’s provision requiring W&T to purchase workers’ compensation insurance for Wood Group employees indicates that the two employers anticipated a borrowed-employee relationship. Given the conflicting evidence, however, we cannot conclude that W&T conclusively established its control of Fredieu. W&T points to several cases to support its position that it controlled Fredieu. In Robertson v. W&T Offshore, Inc., a cook on a W&T platform was held to be W&T’s borrowed employee. See 712 F. Supp. 2d 515, 537 (W.D. La. 2010). Notably there, the cook’s “day-to-day activities were directed by his W&T supervisors.” Id. at 525–26. In Magnon v. Forest Oil Corp., a rig operator was held to be Forest Oil’s borrowed employee. See No. 06-0587, 2007 WL 2736612, at *7 (W.D. La. 2007) (mem. op.). While the operator there argued that he had received all of his skilled training prior to working on the Forest Oil rig, he also admitted “that he received his daily instructions from and was directed by Forest Oil employees.” Id. at *3. In Hotard v. Devon Energy Production Co., a mechanic was held to be Devon’s borrowed employee. See 308 F. App’x 739, 743 (5th Cir. 2009). There, the mechanic was “supervised totally by Devon employees” and testified that “he had no contact with [his original employer] while on the platform.” Id. at 742. In fact, there was only one factor that even conceivably cut in favor of the employee in that case. Id. Finally, in Jackson v. Total E&P USA, Inc., a mechanic was held to be the borrowed employee of Total E&P. See 341 F. App’x 85, 87 (5th Cir. 2009). There, all of the mechanic’s work “was directed by and for the benefit of Total,” and the mechanic’s original employer “was essentially operating as a placement agency.” Id. None of these cases involves evidence of independence as strong as Fredieu’s, including testimony from a W&T vice-president confirming Fredieu’s claim to control his own work on the platform. Again, it was W&T’s burden to prove that Fredieu was its borrowed employee. The primary evidence W&T offers of its control of Fredieu is direction for things “that needed to be done outside of a routine activity” and instructions related to the Job Safety Analysis occurring immediately before his accident. This is insufficient to demonstrate that W&T had the level of “power to control and direct another person in the performance of his work” necessary to convert Fredieu into its borrowed employee. Hebron, 634 F.2d at 247. “‘Co-operation,’ as distinguished from ‘subordination,’ is not enough to create an employment relationship.” Ruiz, 413 F.2d at 313 (citing Standard Oil, 212 U.S. at 226). Even accepting for the sake of argument the idea that Fredieu controlling his own day-to-day work is not relevant to the control analysis, based on the other available evidence, W&T failed to conclusively establish that it did control him. After considering all the Ruiz factors, we conclude that, on the whole, W&T did not establish that Fredieu was its borrowed employee. C. Damages W&T also challenges the legal sufficiency of the evidence to support the jury’s award of $950,000 in damages for Fredieu’s future lost earning capacity. The “test for legal sufficiency must always be whether the evidence at trial would enable reasonable and fair-minded people to reach the verdict under review.” City of Keller v. Wilson, 168 S.W.3d 802, 827 (Tex. 2005). If more than a scintilla of competent evidence supports the judgment, the jury’s verdict must be upheld. Burbage v. Burbage, 447 S.W.3d 249, 259 (Tex. 2014). Fredieu called an economist, Kenneth McCoin, to provide expert witness testimony concerning Fredieu’s future lost earning capacity. McCoin estimated that Fredieu had a pre-injury future earning capacity of $1,611,954 based on earnings of $38,000/year working for Wood Group. He made a low-end calculation of Fredieu’s post-injury future earning capacity of $576,530 based on earnings of $18,000/year. After W&T pointed out that Fredieu had already returned to work for a company called Berry Brothers at a higher hourly rate (though with far fewer hours) than he earned from Wood Group, McCoin adjusted his calculations to estimate a high-end post-injury future earning capacity of $946,530 based on yearly earnings of $28,758 in Fredieu’s new position. These estimates left a range of future lost earning capacity of $665,424 on the low end and $1,035,424 on the high end. The jury ultimately awarded Fredieu $950,000 as damages for future lost earning capacity. In addition to the expert’s calculations, the jury heard evidence about Fredieu’s medical condition, which was relevant to his future earning capacity, especially given his history working physically demanding jobs on offshore rigs. Fredieu has two plates and thirteen screws in his injured arm, is restricted from jobs requiring heavy lifting, and will always have pain and a limited range of motion. Fredieu also testified that, after his injury and prior to working for Berry Brothers, he was out of work for two years. At trial, W&T pointed to the fact that Fredieu returned to work full time after his accident and earned more per hour than he did working for Wood Group. The evidence established that Fredieu was earning $16/hour for regular hours and $24/hour for overtime at Wood Group. After the accident, Berry Brothers paid him $20/hour. By trial, he had received a raise to $23/hour. W&T argues that both of McCoin’s estimates failed to account for Fredieu’s $23/hour wage at the time of trial and are thus unreliable. But Fredieu offered McCoin’s testimony as a “starting point” for the jury to use as a baseline in calculating his future lost earning capacity. The jury was aware of the $23/hour figure and could have taken it into account. W&T also argues that McCoin did not offer testimony of actual lost earning capacity, instead resorting to calculations of the “time-value of money, the probable number of years that Fredieu would work in the future, and other figures used in calculating lost earning capacity.” However, W&T made no objection or motion to exclude McCoin’s testimony. When expert “opinion is admitted in evidence without objection, it may be considered probative evidence even if the basis for the opinion is unreliable.” City of San Antonio v. Pollock, 284 S.W.3d 809, 818 (Tex. 2009). Thus, even if McCoin’s testimony would have been inadmissible as expert testimony if objected to, it was nevertheless some probative evidence of lost future earning capacity for the jury to consider. W&T points to our prior statement that “[i]f [a] plaintiff’s earning capacity is not totally destroyed, but only impaired, the extent of his loss can best be shown by comparing his actual earnings before and after his injury.” McIver v. Gloria, 169 S.W.2d 710, 712 (Tex. 1943). In the same case, however, we also observed that “[i]n a personal injury suit the amount which the plaintiff might have earned in the future is always uncertain, and must be left largely to the sound judgment and discretion of the jury.” Id. We further stated that the plaintiff “is not required to prove the exact amount, but only the facts from which the jury, in the exercise of sound judgment and discretion, can determine the proper amount.” Id. at 712–13. Finally, we included as proper considerations “the jury’s common knowledge and experience and sense of justice.” Id. at 713. While standards of evidentiary review have evolved since 1943, we should not selectively rely on one statement from McIver and ignore all the others. We agree with W&T that evidence of actual earnings at the time of trial is the best evidence of future earning capacity, but it is not the only evidence in an inquiry that looks many years or decades into a person’s future. It is not disputed that Fredieu earned a higher hourly wage working for Berry Brothers after his accident. However, he worked far fewer hours in his new position, and the jury heard testimony that his new position was less lucrative than the salary would indicate. It was 700 miles from his home in Louisiana, and his living expenses were not covered, as they had been working on offshore rigs. Further, the jury was not limited to consideration of evidence concerning Fredieu’s hourly wages in raw numbers; rather, the consideration is one of earning capacity. Reliance Steel & Aluminum Co. v. Sevcik, 267 S.W.3d 867, 872 (Tex. 2008) (“It is of course true that [the plaintiff] was entitled to damages for future earning capacity, not just future earnings.”); McIver, 169 S.W.2d at 713 (“[T]he measure of [the plaintiff's] loss is his capacity to earn, not his actual earnings.”). Given Fredieu’s physical impairment, the work options available to him in the future were limited. The jury heard testimony that Fredieu has been fortunate to find a light-duty job at Berry Brothers through help from relatives and that if he lost that position—in the volatile oil and gas industry—it would be difficult to find a comparable one given his impairment and experience. It was within the “sound judgment and discretion” of the jury to take into account these facts in arriving at a proper damages amount. Id. at 712–13. In evaluating the legal sufficiency of a damages award, we do not lightly reject the judgment of a jury. Houston Unlimited, Inc. Metal Processing v. Mel Acres Ranch, 443 S.W.3d 820, 838 (Tex. 2014). We conclude the jury heard sufficient evidence to support its calculation of lost earning capacity. The amount awarded was not backed by scientifically provable calculations, but few future-earning-capacity awards could be. The jury could certainly have credited W&T’s arguments and awarded a lower amount, but “[i]f the evidence conflicts, it is the province of the jury to determine which evidence to credit.” Id. at 833. In light of the abundance of medical and other evidence provided by witnesses and Fredieu himself, a reasonable and fair-minded jury could have arrived at the number this jury did. Accordingly, we uphold the jury’s award of $950,000 to Fredieu for future lost earning capacity and affirm the court of appeals on this issue. Conclusion We hold that borrowed-employee status is a legal question to be answered by the courts, subject to any subsidiary fact-findings that may be necessary. We also hold that, based on the disputed evidence and the absence of fact-findings in W&T’s favor, W&T failed to prove its borrowed-employee defense. Although we do not agree with all of the court of appeals’ reasoning, we agree with the court of appeals that Fredieu was entitled to judgment on the jury’s findings of liability and damages. For these reasons, the court of appeals’ judgment is affirmed. James D. Blacklock Justice OPINION DELIVERED: June 5, 2020