In this diversity action for damages arising out of asbestos exposure, GAF Corporation appeals the district court's judgment upon a jury verdict in favor of the plaintiff-appellee, James Good. We affirm.

James Good was an insulation worker from 1958 through May, 1980, during which time he developed a tumor in his lung. Part of his lung was surgically removed and medical reports indicated presence of a squamous cancer at the site. Advised by his doctor that a likely association existed between his former occupation as an installer of asbestos-containing products and his lung tumor, Good filed a damage action in 1980 against various asbestos manufacturers and distributors, including GAF.

After recovering from the lung surgery, another tumor was discovered in October, 1982, in Good's small intestine. This tumor was also removed, and medical reports indicated that the intestinal tumor was a seeding or metastasis of the earlier lung tumor. Under this theory, both cancers had a single origin, the lung cancer being primary and seeding the intestinal cancer two years later. This was the theory initially advanced by Good at trial.

GAF's alternative theories were (1) the tumors were not squamous but were instead carcinoid, a low-grade malignancy unrelated to either asbestos or cigarettes, and that the lung cancer was a metastasis of the intestinal cancer, or (2) if the tumors were squamous, the cause was more likely cigarettes than asbestos. Dr. Breitenecker was Good's expert on this issue, and Dr. Auerbach was GAF's.

Less than one month before trial began, Dr. Breitenecker was deposed. The day before, Breitenecker had received Dr. Auerbach's report diagnosing the tumors as carcinoid. In light of this report, Dr. Breitenecker testified at the deposition that he would like to conduct additional tests to test his earlier diagnosis of squamous intestinal cancer. After the new tests were completed, a second deposition of Dr. Breitenecker was conducted a week into the trial. At this time, Dr. Breitenecker stated that his opinion had changed and that he now believed that both the lung and intestinal tumors were primary i.e. neither being the result of the other, and further that, while he still believed the lung cancer to be squamous, he now believed the intestinal cancer to be carcinoid.

A day prior to the second deposition, Dr. Breitenecker telephoned Dr. Eggleston, an expert on lung cancer at Johns Hopkins. According to Dr. Breitenecker, he had previously consulted with Dr. Eggleston on "about a million cases." Dr. Eggleston opined that the cancers were unrelated but that only the lung cancer was squamous. During their conversation, it came out that Dr. Eggleston had been hired by the defense on a consulting basis but that he was not to be called as a witness by GAF. Dr. Eggleston also stated that he knew that Dr. Breitenecker was involved on Good's side of the case. GAF moved to exclude Dr. Breitenecker's testimony on the grounds that his consultation with Dr. Eggleston violated GAF's work product privilege inasmuch as Dr. Eggleston had been retained by GAF as a non-testifying consultant. The court denied the motion to exclude his testimony completely, but ordered Dr. Breitenecker not to mention Dr. Eggleston by name nor to refer to their conversation. His testimony was consistent with that at the second deposition.

After the conclusion of GAF's case, Good called Dr. Eggleston as a rebuttal witness. Over GAF's objections, Dr. Eggleston was permitted to testify. The court reasoned that the advancement of the carcinoid lung defense was a "new theory" which Good should be permitted to rebut with new evidence; the trial judge also expressed concern that the jury should get as much information as possible on a very difficult and technical issue. Dr. Eggleston's testimony was limited to the nature of the lung disease (carcinoid or squamous).

Good also called an economist, Dr. Cobb, to testify about future wage loss. Over GAF's objection, Dr. Cobb was permitted to give his opinion that twenty-five years from now the social security system will not be able to support retirees at the ages of 62 to 65, and therefore more persons currently 40 years old (like Good) will be working longer; he added that the increased working years will also be due to medical advances which will permit more people to work longer.

At the conclusion of its case, GAF offered two exhibits into evidence and, upon Good's hearsay objection, the exhibits were not admitted. Exhibit 5 is an article on work life expectancy by a demographic statistician, published in a Department of Labor magazine. Exhibit 6 contains two tables published by the Census Bureau which described average income at various ages by education and occupation.

The case was submitted to the jury, and a verdict was returned for Good in the amount of $600,000. The trial court granted GAF's motion for credit for prior settlements by the other defendants and, on April 8, 1985, judgment was entered for Good in the amount of $60,000.1 GAF's motion for judgment n.o.v. or for a new trial was denied by order entered on September 24, 1987, and this appeal followed.

GAF contends that the trial court's decision to permit Dr. Eggleston to testify as Good's rebuttal witness was erroneous on the grounds that the testimony should have been excluded due to the relationship between GAF and Dr. Eggleston or, regardless of the relationship, that the testimony should have been excluded as not being proper rebuttal testimony. GAF also contends that the trial court erred in allowing Good's economic expert to state his opinion regarding the future of the social security system. Finally, GAF contends that it was error for the trial court to exclude from evidence the two government reports offered as exhibits. We will discuss each of these contentions in turn.

GAF first characterizes Dr. Eggleston as an expert within the ambit of Fed.R.Civ.P. 26(b)(4)(B). This discovery rule provides that a party may discover opinions of an expert retained or specifically employed by the other party in preparation for trial but who is not expected to be called at trial, only upon "a showing of exceptional circumstances under which it is impractical for the party seeking discovery to obtain facts or opinions on the same subject by other means." GAF's argument goes not to the fact that Dr. Eggleston was called by Good to testify but, rather, to the pretrial conversation between Drs. Breitenecker and Eggleston which led to the decision to call the latter in rebuttal. It is doubtful whether Good could have discovered Dr. Eggleston's opinion through the normal discovery procedures, i.e. upon a showing of exceptional circumstances, but there is no allegation that Dr. Breitenecker's telephone call to Dr. Eggleston constituted discovery, was anything out of the ordinary, or was prompted by anything more than a desire for a knowledgeable second opinion.

According to Professor Moore, the limits on the discovery of non-testifying experts' opinions is based on a "concept of fairness." 4 Moore's Federal Practice p 26.66 (1987). This rule is designed to preclude access to an opposing party's diligent trial preparation. See Advisory Committee Notes, Fed.R.Civ.P. 26(b)(4)(B). In support of its contention that it was error to permit Dr. Eggleston to testify for the plaintiff, GAF relies on Durflinger v. Artiles, 727 F.2d 888 (10th Cir.1984); the cases are distinguishable, however. In Durflinger, the defendants themselves (or, more probably, their attorney) requested and obtained a copy of a report from a non-testifying expert which was prepared for the plaintiff. In upholding the trial court's refusal to permit the defendants to call this expert as their own witness, the Tenth Circuit pointed to the defendant's disregard of the rule and the potential prejudice to plaintiffs' case which the expert's testimony might have presented. The court also noted that under different circumstances, a trial judge might not have to exclude the testimony of a witness who was contacted in violation of Rule 26. Id., at 891. In contrast to this situation, Dr. Breitenecker was neither an opposing party nor its attorney so the opposing party, i.e. Good, did not disregard the rule. Any prejudicial effect which may have been created by virtue of the fact that Dr. Eggleston was originally retained by GAF was avoided by the trial court's insistence that this fact not be made known to the jury. Thus, we hold that, in the context of this case, the contact with Dr. Eggleston did not constitute a violation of Rule 26 and that the court did not err in refusing to exclude his testimony on this basis.

GAF's second contention regarding the admissibility of Dr. Eggleston's testimony is that the content did not qualify as proper rebuttal evidence. GAF argues that Dr. Eggleston's testimony did not counter new facts presented during the defense's case-in-chief but rather that Good only intended this testimony to bolster the earlier testimony of its own expert. GAF characterizes this as an attempt to gain a "last on" effect before the case was submitted to the jury.

The decision to admit this testimony was based primarily on the trial court's belief that the complexity of the medical issues warranted the admission of this additional evidence. Noting that GAF had attempted to prove that Good's lung tumor was carcinoid rather than squamous, the trial court ruled that Dr. Eggleston's testimony would be limited solely to rebutting this defense theory. The trial court believed that the defense's emphasis during trial on one of its two alternative theories provided additional justification to allow this rebuttal evidence. We are unwilling to say that the trial judge abused his discretion in this regard.

GAF's next contention involves the trial court's decision to allow Good's economic expert, Dr. Cobb, to offer his opinion on the future of the social security system. Characterizing this testimony as being impermissibly speculative, GAF asks this Court to find that such evidence was so prejudicial as to constitute reversible error.

The particular portion of Dr. Cobb's testimony to which GAF objects must be examined in light of his entire testimony. The purpose in calling Dr. Cobb as a witness was to prove Good's future earning capacity had he not become disabled. Dr. Cobb opined that people in the coming years will work until later ages for a variety of reasons. One factor which will tend to increase working lives would, according to Dr. Cobb, be improved health. Another factor would be changes in the social security system mandated by economic necessity. Benefits would decrease or the age of eligibility would increase; either way, these changes would act to increase the number of years in the average working life and thereby increase the average future earning capacity.

In one sense, of course, any opinion on matters twenty-five years in the future is based to some degree on speculation. Trial courts, however, are endowed with a great deal of latitude in determining where to draw the line between permissible and impermissible speculation. See e.g. United States v. Portsmouth Paving Corp., 694 F.2d 312, 323 (4th Cir.1982). We note that GAF cross-examined Dr. Cobb regarding the basis of his opinion on the future of the social security system; although it chose not to do so, GAF was free to counter Dr. Cobb's testimony with expert testimony of its own. Dr. Cobb's testimony does not strike us as beyond the limits of permissible speculative opinion evidence, and we hold that it was properly admitted.

GAF's final contention is that the trial court erred in denying its motion to have two exhibits introduced into evidence. The court sustained Good's hearsay objection, but GAF contends that the proffered documents were within the public records exception found at Rule 803(8) of the Federal Rules of Evidence. Mindful of our previous rulings which uphold a liberal construction of Rule 803(8), we decline to decide whether the exclusion of the exhibit evidence was erroneous because we feel the better course is to uphold the trial court's ruling on the basis of harmless error. See, e.g., Ellis v. International Playtex, Inc., 745 F.2d 292 (4th Cir.1984).

The harmless error rule requires the trial court to disregard any error in the exclusion or admission of evidence "which does not affect the substantial rights of the parties." Fed.R.Civ.P. 61. The same policy is applicable to appellate courts. 28 U.S.C. Sec. 2111.2 As the Supreme Court noted in McDonough Power Equipment, Inc. v. Greenwood, 464 U.S. 548, 553 (1984), appellate courts should "ignore errors which do not affect the essential fairness of the trial." Assuming that it was error to exclude GAF's exhibits, we hold that such error was harmless.

GAF's lawyers made extensive use of the two exhibits during the cross-examination of Good's economic expert, Dr. Cobb. Dr. Cobb admitted that he was familiar with the article marked as GAF's exhibit No. 5 and that a government study cited by the author estimated a working life expectancy for someone of Good's age to be approximately 60.8 years. Similarly, Dr. Cobb was questioned regarding certain statistics contained in GAF's Exhibit No. 6, a report prepared by the U.S. Bureau of Census. During cross-examination, GAF's lawyer was able to elicit from Dr. Cobb the observation that one table in the exhibits reported that the mean income for males of Good's age and educational background was between $13,712 and $16,417. We are unable to imagine what else GAF hoped to have the jury glean from these documents, and we certainly cannot understand how the exclusion of these documents affected the "essential fairness" of the trial.

For all the foregoing reasons, the judgment of the district court is affirmed.

AFFIRMED.