BEAM, Circuit Judge.
FMC Corporation appeals a judgment in favor of Todd Gander for $2,000,000, entered on July 13, 1988, following a jury trial. Gander’s theories of recovery for personal injuries sounded in both strict liability in tort and negligence, and the jury found in favor of Gander on both claims. The jury also assessed Gander’s comparative fault on the negligence claim at 90%. On appeal, FMC argues that the verdict form was confusing and misstated the law, [1375]*1375and that the district court should have allowed FMC to cross-examine Gander’s witnesses about the effects of income tax assessments on Gander’s lost income. While we have some reservations about the verdict form, we affirm.
I. BACKGROUND
Gander lost his right arm at the elbow on February 23, 1985, while working for An-heuser-Busch in St. Louis. Gander worked as a coal conveyor operator, and made $34,-560 plus benefits in 1984. At the time of trial in 1986, the average pay for coal conveyor operators was $47,891 plus benefits.
As a coal conveyor operator, Gander’s work entailed shepherding coal from its entry into the plant to the boilers where it was converted to electrical energy. The conveyor belts which transported the coal frequently would ride to one side of their drive drums as the result of coal dust accumulating on the drums. When the convey- or belt rode to one side, it would often climb the wall, causing coal spills.
On February 23, 1985, Gander was working near a head pulley drum which had become clogged with coal, and which was partly responsible for a major coal spill. Gander had been taught to clean the drum with a fox tail brush, without shutting off the system. Gander thus proceeded to clean the drum with the fox tail brush, but the rolling drum caught the fox tail brush, pulling Gander’s forearm into the device before he was able to let go. The rolling drum literally lifted Gander off the ground, and his arm was severed at the elbow when it came into contact with a shear guard. Fellow workers took Gander to Barnes Hospital where his arm was reattached.
Although Gander’s surgery was successful to the extent that he still has his right arm, he testified at trial that the arm is essentially useless. “It’s just like an ornament hanging on, it’s no good.” Trial Transcript, vol. 1, at 122. He has only limited movement in his shoulder and must wear a brace to keep his wrist straight. It otherwise flops limply. Gander has no feeling in his fingers, has lost his fingernails, and must wear a glove on his hand to prevent swelling. Gander’s useless right arm causes him problems with minor things, like buttoning the top of his shirt or tying his shoe laces; Gander testified that these problems cause him to be frustrated and irritable. Moreover, attempts at vocational rehabilitation were unsuccessful, and Gander was discharged from Anheuser-Busch on August 31, 1987. He now does some volunteer work at a local hospital. The trial testimony indicated that the injury has enormously changed Gander’s life.
As indicated, the case was tried to a jury on theories of strict products liability and negligence. The single verdict form submitted to the jury contained reference to both theories. Part I required a verdict on the strict liability claim. Part II required the jury to assess the comparative fault of the plaintiff and the defendant in the event that it found for the plaintiff on the negligence claim. Part III required the jury to assess total damages. The jury found for Gander on both theories, assessed Gander’s fault at 90%, and found total damages to be $2,000,000.
After the jury was dismissed, however, the district court expressed doubt about whether the jury verdict was for $2,000,000 or for $200,000; i.e., $2,000,000 total damages reduced by the 90% fault assessed to Gander. The district court initially entered judgment for $200,000. On Gander’s motion to amend the judgment, however, the court decided that Gander was entitled to $2,000,000 on the strict liability claim, which could not be reduced by Gander’s comparative fault under Missouri law. This appeal followed.
II. DISCUSSION
A. The Verdict Form
FMC argues on appeal that the district court erred both in submitting to the jury Gander’s verdict form, which it argues was confusing and inaccurate, and in amending judgment. While the arguments are closely related and both ultimately turn on whether the verdict form correctly stated Missouri law, we treat the arguments separately.
[1376]*1376The verdict form was submitted by Gander, and was taken from the Missouri Approved Jury Instructions. MAI 35.15 illustration (3d ed. Supp.1989) (the form is indicated as Verdict A, and is taken from MAI 36.01 and 37.07). The form used is approved, at least as an illustration, by the Missouri Supreme Court for the submission of both strict liability and negligence claims, and was new at the time of trial. Brief for Appellant at 9. It contains three parts,1 the first dealing with the strict liability claim, the second with comparative fault on the negligence claim, and the third with total damages. The difficulty the district court had, and the arguable error in the form, stems from the closing note, which explains the reduction in damages for plaintiffs comparative fault on the negligence claim.
NOTE: If you assessed a percentage of fault to any defendant on plaintiffs claim for personal injury based on negligence, the judge will reduce the total amount of plaintiffs damages by any percentage of fault you assess to plaintiff.
After the jury had been dismissed, FMC remarked that the verdict appeared to be $200,000; i.e., $2,000,000 total damages reduced by the 90% fault assessed to Gander. Trial Transcript, vol. 3, at 270. Gander argued that the verdict was for $2,000,000, since the jury found for Gander on the strict liability claim, which could not be reduced by Gander’s comparative fault. Id. at 272. The district court replied: “Not based on the note on part three.” Id. After quoting the note, the court continued: “This is directing me to reduce that by 90 percent.” Id. at 274. The district court then entered judgment for $200,000.
Upon Gander’s motion to amend, however, the district court decided that it had improperly reduced the total damages by the percentage of comparative fault assessed to Gander. “Inasmuch as this is a case in which comparative negligence principles apply only to the jury’s determination on the negligence claim and not to the strict liability claim, the jury’s findings as reflected on the verdict form entitle plaintiff to a judgment in the amount of Two Million Dollars ($2,000,000).” Memorandum and Order, No. 87-1155C(6), July 13, 1988, at 2.
To the extent that FMC argues that the district court erred in amending its judgment, we disagree. “The decision to grant or deny a Rule 59 motion is committed to the sound discretion of the trial court.” A.W. v. Northwest R-1 School Dist., 813 F.2d 158, 165 (8th Cir.), cert. denied, 484 U.S. 847, 108 S.Ct. 144, 98 L.Ed.2d 100 (1987); Slater v. KFC Corp., 621 F.2d 932, 939 (8th Cir.1980). Under Missouri law, which we are bound to apply [1377]*1377in this diversity ease, see Walker v. Paccar, Inc., 802 F.2d 1053, 1055 (8th Cir.1986), the district court was correct in amending the judgment. Missouri law is clear that a verdict resting on a strict liability claim cannot be reduced by a plaintiffs comparative fault. In Lippard v. Houdaille Indus., 715 S.W.2d 491 (Mo.1986), the Missouri Supreme Court held that “the plaintiffs contributory negligence is not at issue in a products liability case.” Id. at 493. The court explained that negligence on the part of plaintiff should neither prohibit nor reduce plaintiffs recovery on a strict liability claim. “We adhere to the view that distributors of ‘defective products unreasonably dangerous’ should pay damages for injuries caused by the products, without reduction because a plaintiff may have been guilty of a degree of carelessness.” Id. at 494. As defense counsel admitted at oral argument, the judgment of $200,000 under the strict liability theory was clearly incorrect as a matter of Missouri law. Therefore, we find no abuse of discretion by the district court.
This does not, however, end the matter. FMC’s argument is better stated in terms of the verdict form’s correctness. That is, the fact that the amended judgment properly reflected Missouri law does not mean that the verdict form itself did not misstate Missouri law or unduly confuse the jury. Thus, FMC argues that the form was incorrect as a matter of law, or, in the alternative, that it was so confusing that the intent of the jury is unclear. The jury may, indeed, have been misled by the closing note, which could be read (as first read by the district court) to apply comparative fault to the strict liability claim as well as to the negligence claim. On its face, the note is not specifically limited to the negligence claim. Nor does the note specifically apply to the strict liability claim. Because of this ambiguity, the jury, according to FMC, may have been led to conclude that total damages would be reduced by any percentage of fault assessed to Gander, even though it found in Gander’s favor on the strict liability claim. Hence, FMC argues, the jury probably intended that Gander receive only $200,000, and thus assessed total damages of $2,000,000 because it thought that total damages would be reduced by 90%. We must, therefore, consider whether the verdict form is legally incorrect or unduly misleading.
While it is a close question, we do not think that the verdict form misstates Missouri law. FMC agreed at oral argument that the difficulty with the form stems from the closing note, which, as indicated, can be read to require a reduction, based on plaintiff’s comparative fault on the negligence claim, in the strict liability claim. Although the note leaves open this possibility, it does not, as the district court initially thought, require it. Because of its grammatical structure and location, the note is somewhat misleading. The note can be read to indicate a connection between a finding of contributory fault on the negligence claim and a reduction in total damages. If the jury assesses a percentage of fault to any defendant, then “the judge will reduce ” total damages accordingly. The reduction in damages thus can be caused by the jury assessing fault, regardless of any finding on the strict liability claim. While this is a possible reading of the note, it is not a required reading.
However, as a matter of law, the note is not incorrect; it is simply unclear. Given that Missouri law is clear that the judge cannot apply comparative negligence principles to the strict liability claim, the strict liability claim will, in fact, be unaffected by any percentage of fault assessed to plaintiff on the negligence claim. The note does not require otherwise, and thus is not legally incorrect. It does not misstate Missouri law.
The verdict form is, however, at least potentially confusing to the jury.2 [1378]*1378We must, therefore, consider whether the form, taken together with the instructions to the jury, is so misleading or confusing that the jury verdict cannot stand. In determining whether the verdict form is confusing, we must consider it in light of the instructions given. See United States v. Hines, 728 F.2d 421, 427 (10th Cir.), cert. denied, 467 U.S. 1246, 104 S.Ct. 3523, 82 L.Ed.2d 831 (1984). It is enough if the “charge as a whole ... state[s] the governing law fairly; technical imperfections or a lack of absolute clarity will not render the instructions erroneous.” Toro Co. v. R & R Products Co., 787 F.2d 1208, 1215 (8th Cir.1986). We find that the jury was correctly and adequately instructed on Missouri law, and, therefore, that the verdict form, not clearly wrong, also is not so confusing that the jury verdict must be upset.
The important inquiry is whether the jury was properly instructed that damages were to be calculated regardless of any fault assessed to plaintiff on the negligence claim. Initially, we note that the verdict form itself is clear and correct on this matter. In Part III, in which the jury finds total damages, the form states: “We, the jury, find the total amount of plaintiffs damages disregarding any fault on the ■part of plaintiff to be_” (emphasis added). This instruction that comparative fault is not to be considered in calculating total damages is much clearer than is any inference from the concluding note that might lead the jury to inflate total damages in anticipation of a reduction for plaintiff’s comparative fault on the negligence claim. Instruction number 18 further provided that, if the jury found in favor of Gander on the strict liability claim, or assessed a percentage of fault to FMC on the negligence claim, then
disregarding any fault on the part of plaintiff, you must determine the total amount of his damages to be such sum as will fairly and justly compensate him for any damages you believe he sustained and is reasonably certain to sustain in the future as a direct result of the occurrence mentioned in the evidence.
Moreover, the jury was specifically instructed, through instruction number 13, that it should calculate total damages regardless of plaintiff’s fault. “In determining the total amount of plaintiff’s damages, you must not reduce such damages by any percentage of fault you may assess to plaintiff.” Instruction number 13, see also MAI 37.03 (3d ed. 1989 Supp.). Thus, the district court correctly instructed the jury that its function was not to determine the amount of plaintiff’s recovery, but to determine the amount of plaintiff’s total damages. Any efforts by the jury to calculate damages in terms of plaintiff’s recovery would be contrary to the court’s instructions.
Not only was the jury properly instructed on its role in calculating damages, but the district court also properly instructed the jury that comparative fault principles do not apply to strict liability claims. Instruction number 13 provided that: “The judge will compute any recovery on plaintiffs claim for personal injury based on negligence by reducing the amount you find as plaintiff’s total damages by any percentage of fault you assess to plaintiff.” Instruction number 13, see also MAI 35.15 illustration (3d ed. Supp.1989) (emphasis added).3 Thus, while the verdict form may be unclear about whether its concluding note applies to the strict liability claim, the instructions were quite clear that the jury finding on comparative fault does not so apply. As a whole, the instructions correctly and specifically instructed the jury about the ambiguities the form may have created. Given this correct and explicit [1379]*1379instruction, it is mere speculation that the jury calculated total damages at $2,000,000 in anticipation of a reduction for plaintiffs comparative fault on the negligence claim.
Moreover, mere speculation that a jury verdict may have been based on the jury’s own misunderstanding of the law, even though properly instructed, is an insufficient basis on which to upset a jury verdict. “It is well settled that a jury’s misunderstanding of testimony, misapprehension of law, errors in computation or improper methods of computation, unsound reasoning or other improper motives cannot be used to impeach a verdict.” Chicago, Rock Island & Pacific R.R. v. Speth, 404 F.2d 291, 295 (8th Cir.1968). In Speth, the jury assessed plaintiff’s contributory negligence at 40% and awarded $16,000 in damages. The district court, sua sponte, asked the jury whether its damage award was net or gross. The jury responded that its calculation was a net figure, compensating for plaintiff’s contributory negligence. The jury thus intended for plaintiff to recover $16,000. When the district court sent the jury back to recalculate damages, it returned with a figure of $40,000. On appeal, this court reversed and remanded for a new trial on the issue of damages. Id. at 296. This is not to say, however, that when a jury verdict “on its face shows a clear disregard for the court’s instructions,” id. at 295, it cannot be corrected. But given correct instruction on the law and no clear disregard for that instruction on the face of the verdict, a jury verdict must remain immune from questioning by the district court and from speculation by an appellate court that the verdict may be based on a misunderstanding of the law. Whether the jury misunderstood a correct instruction of the law is an improper subject for mere speculation. See 6A J. Moore, Moore’s Federal Practice 1159.08[4], at 59-115 to 116 (1989) (“a verdict cannot be upset by speculation”); id. at 59-127 (“A mere suspicion, however, that the jury has not followed the court’s instructions is not sufficient, since that would make the verdict too vulnerable and would needlessly prolong litigation.”); Peveto v. Sears, Roebuck & Co., 807 F.2d 486, 490 (5th Cir.1987) (“Whether or not the jury misunderstood the charge of the court is not a question to be reexamined after the verdict has been rendered.”) (citation omitted).
Our holding finds support as well in the cases which, based on Federal Rule of Evidence 606(b), forbid improper impeachment of a jury verdict. While no improper impeachment occurred in this case, Rule 606(b) establishes that it would have been improper to inquire of the jurors what they really intended by their verdict. In Karl v. Burlington Northern R. Co., 880 F.2d 68 (8th Cir.1989), the jury returned a verdict in a personal injury action in favor of plaintiff. Id. at 69. The jury found plaintiff to be 75% at fault, id. at 70 n. 3, and assessed damages at $273,750. The jury was instructed to assess damages “without taking into consideration any reduction of the plaintiff’s claim due to her own negligence.” Id. at 70 n. 4. The district court, however, called the jury foreman at the request of plaintiff’s counsel, to determine whether the jury calculated damages without regard for plaintiff's contributory negligence. Id. at 71. The jury foreman explained that $273,750 was 25% of what plaintiff asked for, and was what the jury intended plaintiff to recover. Id. The district court then amended judgment to reflect actual damages of $1,095,000. Id. at 72.
This court reversed the amended judgment. Relying on Rule 606(b), which provides that a juror may not testify about the jury’s thought processes during deliberations for purposes of impeaching the verdict, we found the amendment to be improper. Id. at 73-74. The impeaching testimony concerned “how the jury interpreted the court's instructions, and concerns the jurors’ ‘mental processes.’ ” Id. at 74. The testimony, therefore, violated Rule 606(b). Given that “there is no other indication that the jury’s first verdict was deficient,” id. at 75, this court remanded to reinstate the first verdict.
Similarly, while no impeachment of the jury verdict took place in this case, any inquiry by the district court would have [1380]*1380been improper. Nor is there any “other indication that the jury’s verdict was deficient.” We are, therefore, unwilling to speculate that the jury, even though properly instructed, misunderstood the law and, therefore, did not intend to award plaintiff damages of $2,000,000. See also Scogin v. Century Fitness, Inc., 780 F.2d 1316, 1320 (8th Cir.1985); Peveto, 807 F.2d at 489.
Finally, we find it persuasive that the evidence of damages in this case is sufficient to support the jury’s calculation of damages of $2,000,000. The uncontradicted figures for lost wages, past and future, provided by plaintiff’s expert witness, Viscusi, establish damages in excess of $2,000,000. Viscusi testified that at the contract wage for an oiler (all AnheuserBusch oilers were actually paid more), Gander’s lost wages for the next twenty seven years would be $1,395,711. Trial Transcript, vol. 1, at 316. Given Gander’s substantial earnings growth of 15% for the four years prior to the accident, id. at 307, Viscusi called this figure “ludicrous.” Id. at 317. Rather, Gander’s lost wages were more accurately reflected by considering wages for the average oiler, and would be closer to $1,693,000. Id. at 318. Viscusi also gave lost wage scenarios which calculated Gander’s wage growth at better than the rate of interest, and gave figures ranging from $1,808,714 to $2,217,761. Id. at 320-22. Gander’s counsel used these figures, together with past wage losses, medical bills and pain and suffering, to argue for $3,000,000 in damages. Id. vol. 3, at 234-35.4
Nor were these figures at all excessive in light of Gander’s injury. Gander’s physician testified that Gander’s right arm is almost useless; that he has no use of his fingers, which frequently swell and develop ulcers; that he has permanent, burning pain in his thumb, index and middle fingers, which are otherwise numb; that he must wear a splint to support his wrist and a glove to keep down the swelling in his hand; that his arm is subject to injury and heals slowly; that the swelling in his arm is likely permanent; and that Gander’s arm is 95% permanently disabled. Id. vol. 2, at 34-59.
Gander’s psychiatrist also testified that the injury has caused substantial psychological problems. When Gander first contacted the psychiatrist, he was severely depressed and suicidal, and had to be hospitalized for treatment. Id. at 75. The psychiatrist still sees Gander monthly, and testified that his symptoms are in a constant state of flux. Id. at 80. While Gander is no longer suicidal, the scars on his body from skin grafts, nerve and vein transplants, and muscle harvesting are devastating to Gander’s self-image. Id. at 77. Gander is also easily frustrated and irritable, suffers from insomnia, a decreased interest in intimacy with his wife, feelings of hopelessness and worthlessness, and generally has a poor outlook for the future. Id. at 90. Moreover, the psychiatrist testified that, on the basis of psychology alone, Gander should not return to work, where he would likely be frustrated and perhaps disposed to again think of suicide. Id. at 84. The psychiatrist testified that Gander’s long range outlook is poor. Id.
Thus, given the substantial evidence supporting the jury’s calculation of damages at $2,000,000, the correct instruction of the law, and mere speculation that the jury may have intended to award Gander only $200,000, we cannot set aside the jury verdict on the ground that the verdict form was unclear.
B. Income Taxes
FMC next argues that the district court erred in not permitting cross-examination of Gander’s witnesses about the effect of income taxes on Gander’s lost income. FMC first attempted to cross-examine Vis-cusi, the economist who testified about Gander’s lost wages. FMC asked Viscusi whether the figures he was using were in terms of gross income. Counsel then asked whether Gander would actually get those gross dollars, to which Gander objected. Trial Transcript, vol. 1, at 340-41. At [1381]*1381the bench, Gander said that under Missouri law, income tax assessments were inappropriate subjects for cross-examination, and that any contrary holdings from FELA cases could not apply in this diversity case. Id. at 342. FMC suggested that it should be able to cross-examine the economist, but when asked by the court whether the law was clear that such cross-examination was permissible, FMC counsel responded: “I think I can.” Id. at 343. The district court allowed FMC to cross-examine on income taxes, but also said that it would grant a mistrial if Gander were able to show that such cross-examination was improper. Id. at 344. FMC made no further efforts to cross-examine Viscusi about income taxes.
The second exchange over income taxes occurred at the bench, when FMC presented to the court Nesselrode v. Executive Beechcraft, Inc., 707 S.W.2d 371 (Mo.1986), which FMC said permitted cross-examination on income taxes. Trial Transcript, vol. 2, at 2-3. The district court noted, however, that Nesselrode made the law no clearer, and thus adhered to its original ruling. Id. at 4.
FMC then attempted to cross-examine Crawford, Gander’s witness from the union. Counsel asked Crawford, who testified about wages earned by coal conveyor operators at Anheuser-Busch, whether his figures were net or gross. Gander objected on the basis of the previous discussions, and this time the objection was explicitly sustained. Id. at 70-71.5
Finally, FMC tendered an instruction on reduction to present value, which it then linked to the issue of whether the total award was subject to income taxation. Id. vol. 3, at 214-15. Gander objected to the instruction and the court agreed, thus precluding FMC from arguing the taxability of the award to the jury. Id. at 215.6 It is somewhat unclear from the objections and the discussions at trial whether FMC appeals from the district court’s rulings on cross-examination or from the district court’s refusal to instruct the jury that the judgment was not subject to income taxation. In either case, however, Missouri law is clear that the trial court did not err in its rulings.
We must first consider the appropriate governing law, for FMC argues in its brief that these issues are controlled by Norfolk & Western Ry. v. Liepelt, 444 U.S. 490, 100 S.Ct. 755, 62 L.Ed.2d 689 (1980) and Jones & Laughlin Steel Corp. v. Pfeifer, 462 U.S. 523, 103 S.Ct. 2541, 76 L.Ed.2d 768 (1983). In Liepelt, an FELA action, the Supreme Court considered whether the jury should receive both evidence of the effect of income taxes on lost wages and an instruction on the taxability of a final award. As to the effects of taxes on a lost income stream, the Supreme Court held that such evidence was neither too speculative nor complex for a modern jury to understand, and thus was proper to establish after-tax earnings. Liepelt, 444 U.S. at 494, 100 S.Ct. at 757-58. The Supreme Court also held that it was error to refuse to give an instruction that the final award was not subject to income tax. Id. at 498, 100 S.Ct. at 759-60.7 FMC argues that [1382]*1382these cases establish that the district court erred in its rulings.8
These cases, however, are not controlling. We have specifically held, in Adams v. Fuqua Indus., 820 F.2d 271 (8th Cir.1987), that Liepelt does not control in non-FELA cases. In Adams, appellant argued that the rationale of Liepelt transcends the FELA context and that it should, therefore, apply in diversity actions. Id. at 276. This court rejected the argument, and, in considering an instruction on the taxability of an award, held that: “Whether to give or withhold a taxability instruction is a question of state law, which we are bound to follow.” Id. at 277. More generally, the law is perfectly clear that “when a tort action is brought in federal court pursuant to diversity jurisdiction, basing liability on state law, the court must apply state law in regard to availability and computation of damages.” Id. (quoting Losey v. North American Philips Consumer Electronics Corp., 792 F.2d 58, 61-62 (6th Cir.1986)). Moreover, this court in Adams specifically rejected a Seventh Circuit case9 which applied Liepelt in a diversity case. We defer instead to “Missouri’s right to create its own law free from federal interference.” Adams, 820 F.2d at 278. Thus, we apply Missouri law.
Missouri law is clear that it is not error for a court to refuse to give an instruction that an award is not subject to income taxation. At one time, Missouri law favored such an instruction. In Dempsey v. Thompson, 363 Mo. 339, 251 S.W.2d 42 (1952), an FELA case, defendant argued error in the court’s refusal to give an instruction that no income tax would be assessed on the judgment. Id. 251 S.W.2d at 43-44. The Missouri Supreme Court agreed, finding no reason not to give the instruction, especially given the danger that the jury could incorrectly overcompensate for taxes on the judgment. It was, therefore, error to refuse the instruction. Id. at 45. More recent Missouri cases, however, establish that Dempsey has been superseded by Missouri Approved Jury Instructions (MAI) outside of the FELA context, and that it is not error to refuse such an instruction.
The Missouri Court of Appeals reviewed several decisions on this issue in Tennis v. General Motors Corp., 625 S.W.2d 218 (Mo.Ct.App.1981). Noting that Dempsey was an FELA case and was decided before the Missouri Approved Jury Instructions were approved by the Missouri Supreme Court, the court explained its ruling in Senter v. Ferguson, 486 S.W.2d 644 (Mo.Ct.App.1972), in which the court held that it was error to give the same instruction given in Dempsey. In Tennis, the court explained that Senter was a post-MAI case, and that MAI 4.01, not Dempsey, controlled the instruction issue in a non-FELA case. MAI 4.01, the only permissible and authorized damage instruction in a non-FELA case, explicitly did not incorporate the Dempsey FELA instruction. Thus, to give the instruction would be error, since the instruction is “clearly an addition to or ‘modification’ of MAI 4.01.” Tennis, 625 S.W.2d at 226. Moreover, the second edition of MAI did not include the Dempsey instruction even in its damage instructions for FELA cases, MAI 8.01 and 8.02. Following the Supreme Court decision in Liep-elt, however, both FELA instructions were modified in the third edition to contain the charge that “ ‘any award you may make is not subject to income tax.’ ” Id. at 227. “It seems worthy of note that although FELA instructions MAI 8.01 and 8.02 were changed in the 3rd Edition of MAI in compliance with Norfolk, MAI [4.01] has not been altered nor amended to include a direction that any awarded damages are ‘not subject to income tax.’ ” Id. Thus, in Tennis, the Missouri Court of Appeals found [1383]*1383no error in the lower court’s refusal to give the instruction in a non-FELA case.
Again, in Kenton v. Hyatt Hotels Corp., 693 S.W.2d 83 (Mo.1985), appellant argued error in the court’s refusal to give a similar instruction. Id. at 96. The Missouri Supreme Court reviewed the law, and affirmed the lower court. “The trial court here cannot be convicted of error, requiring a new trial, in following the established law of this state.” Id. at 96.
Thus, to the extent that FMC presented the court with an instruction that the judgment would not be subject to income taxation, we find no error in the district court’s refusal to give such an instruction.
Missouri law is somewhat less clear, however, regarding the admissibility of evidence relating to the effects of income tax on a lost-income stream. The controlling case appears to be Dempsey. In Dempsey, defendant argued on appeal that the trial court had erred in refusing to permit defendant to cross-examine plaintiff’s witness about the effects of income tax on a personal injury award. Dempsey, 251 S.W.2d at 43. The Missouri Supreme Court held that:
The trial court did not err either in refusing to permit defendant to cross-examine plaintiff’s actuarial witness relative to income tax liability or in refusing to permit defendant to argue to the jury that in arriving at the amount of its award it should consider only the amount of future earnings lost to plaintiff after deduction of income taxes for which he would have been liable had he continued his employment without injury.
Id. at 45. While it is unclear as to whether FMC’s attempted cross-examination concerned Gander’s tax liability on the award or the effects of income taxes on lost income, the Missouri court’s reference to defendant’s proposed argument is exactly on point. This is precisely what FMC attempted to do in this case, and it is this point of law about which FMC admitted to the district court that it was uncertain. To the extent that Dempsey is the controlling law, the district court did not err in refusing FMC’s proposed cross-examination.
The issue is somewhat clouded, however, by Nesselrode, 707 S.W.2d 371. FMC argues that in Nesselrode the Missouri Supreme Court explicitly found the sort of cross-examination that FMC attempted to be permissible. By implication, FMC argues that Nesselrode silently overrules the precise holding of Dempsey which found no error in refusing exactly the sort of cross-examination FMC attempted. We do not agree that Nesselrode controls this case.
Indeed, whether a defendant can cross-examine about the effects of income taxes on plaintiff’s lost income stream was not even at issue in Nesselrode. Rather, the context in which the court makes the reference to cross-examination on which FMC relies concerns a dispute about the reduction of lost income to present value. Both parties in Nesselrode argued error in failing to establish the present value of plaintiff’s lost income. Specifically, defendant argued that the court should not have allowed the jury to consider a table indicating plaintiff’s lost income, since the table contained no reductions to present value. Id. at 385-86. The Missouri Supreme Court characterized the issue even more narrowly. It found the issue to be whether defendant could object to the trial court’s ruling on plaintiff's presentation of lost income when defendant itself failed to present any evidence of present value. Id. at 387.
In this context, the court mentioned in passing that while defendant did not challenge plaintiff’s presentation of present value, defendant did challenge plaintiff’s claim for damages. “This was done by presenting evidence in connection with decedent’s history of health problems, questioning the three daughters’ monetary reliance on decedent, and by pointing out through cross-examination that decedent’s projected stream of lost-income does not take into consideration his income tax obligations.’’ Id. at 388 (emphasis added). We do not believe that this passing reference to cross-examination can be interpreted in such a way as to overrule [1384]*1384Dempsey. The propriety of the cross-examination was not at issue in Nesselrode and was not addressed by the court. Nor, apparently, was the cross-examination in Nesselrode the subject of an objection by plaintiffs counsel at trial. This language in Nesselrode, then, is just a mere, isolated reference in passing to the sort of cross-examination attempted by FMC. This same sort of cross-examination, by contrast, was specifically addressed in Dempsey. FMC presented no persuasive law to the district court in support of such cross-examination, and it likewise presents none here. Thus, we find no error in the district court’s ruling.
III. CONCLUSION
We have considered FMC’s other arguments on appeal and find them to be without merit. For the reasons stated, we affirm the judgment of the district court.