552 F. 2d 1061 - Renfro Hosiery Mills Company v. National Cash Register Company

Norwood Robinson, George Little, Jr., and Steven E. Philo, Winston Salem, N. C. (Hudson, Petree, Stockton, Stockton & Robinson, Winston Salem, N. C.), for appellant.

John R. Surratt, Winston Salem, N. C. (Surratt & Early, Winston Salem, N. C.), for appellee.

Before BUTZNER, Circuit Judge, BRYAN, Senior Circuit Judge, and WATKINS, Senior District Judge.*

The instant appeal arises from an action by Appellee NCR Corporation against Appellant Renfro Corporation1 in which NCR, seeking liquidated damages, alleged that Renfro had breached a contract to lease NCR's computer equipment. Renfro counterclaimed, alleging breach of express and implied warranties, negligence, and fraudulent misrepresentations by NCR. A directed verdict was granted on the issue of negligence, and the case was submitted to the jury on the remaining issues. The jury found that Renfro had breached the lease agreement but awarded no damages, and made no finding with respect to the issues raised in the counterclaim.2 Renfro has now appealed, assigning as error various rulings by the trial judge on evidence offered by Renfro, some of which was excluded and some of which was admitted for certain purposes only.

* Prior to 1969, Renfro, a North Carolina sock manufacturing company, had used manual data gathering techniques to keep track of sales, inventory, production, deliveries, billing, and general accounting. In July 1968, Renfro began negotiations with NCR regarding the possible leasing of automatic data processing equipment. Interest centered on the Century 100 Series, a new computer system then being developed by NCR which had not yet been installed in the field. It appears from the record that Renfro was advised by NCR that the computer was new to the market. Tr. 1131, 1883. Despite the fact that the Century 100 was in fact developmental at the time the negotiations were being carried on, NCR made repeated oral and written representations to Renfro to the effect that the Century 100 was "reliable" or would give "increased reliability." The statements were extremely general, however; reliability was not defined in terms of what specifically might be expected of the equipment.3 Furthermore, it appears that no reference was made to testing then being conducted by NCR on the prototype of the Century 100 or, later, on the first production model.4

In September 1968, the parties executed a contract by which Renfro agreed to lease the Century 100 Series basic system and certain peripheral equipment for five years at a stated monthly rental. This contract was superseded by another agreement executed on March 14, 1969, which essentially repeated the terms of the September contract but which contained certain changes in equipment specifications. Under the terms of the March agreement, Renfro was to pay a monthly charge of $3,515 to NCR. NCR warranted that "the Equipment will be in good working order," disclaiming all other warranties, and agreed to furnish maintenance service for the computer system. The lease further provided that the customer might terminate the arrangement short of the five-year term but that, in such a case, NCR could levy a termination charge. In separate writings, the parties agreed that no rental would be due until both parties were satisfied with the installation and operation of the computer at Renfro's site.

Installation of the computer system began in early November 1969. From November 21, 1969, through December 11 or 12, the computer system was tested; and, on December 12, Renfro informed NCR that it was satisfied with the operation of the computer, although not all of the applications for which the computer was intended had then been fully implemented. Tr. 437-438. Accordingly, from December 11 until termination of the lease, Renfro was billed for rental according to the monthly schedule.

During the first few months of operation, Renfro maintained a "parallel" system of data collection; that is, the old manual system was maintained as a check on the operation of the computer system, at least to some extent. However, on May 15, 1970, Renfro decided to abandon the manual system entirely and "go live" on the computer.

At first, this operation appeared to be satisfactory. In June, however, Renfro claims to have experienced serious problems relating mainly to customer service. It is alleged that these problems were caused by the unreliability of the computer system and the inability of NCR to keep it in repair. In August 1970, Renfro informed NCR that the lease was terminated and directed NCR to remove its equipment. In Renfro's view, it had suffered significant loss of business due to the malfunctioning of the NCR computer, along with loss of good will and consequential damages. NCR did not let the matter rest, however, and insisted upon payment of the termination charge prescribed by the lease as liquidated damages. Renfro refused to pay these charges.

NCR then filed suit, claiming that it was entitled to recover the termination charge prescribed by the lease as liquidated damages. It was asserted that NCR had complied with its obligation to supply the equipment specified in the lease and that the equipment had been maintained in good working order as warranted.

In response, Renfro posed numerous affirmative defenses not relevant to the instant appeal and, in addition, entered a counterclaim for very substantial damages allegedly suffered because of the alleged disastrously poor performance and maintenance record of the NCR computer system. The counterclaim went to the jury on two issues. The first of these was Renfro's contention that repeated malfunctions of the computer and the inability of NCR to maintain the computer in an "up" condition amounted to a breach of express and implied warranties; Renfro averred, of course, that the disclaimer of warranties in the lease was ineffective. The other was the claim that NCR had made repeated representations and assurances to Renfro that the Century 100 was a "reliable" computer system, when in fact NCR was aware that this was not the case, and that Renfro had relied on these assurances to its detriment. As stated above, the jury found that Renfro had breached the lease but awarded no damages and made no finding as to the counterclaim.

Several adverse rulings on evidence are challenged in this appeal. These are discussed individually in the subsequent sections of this opinion.

Appellant complains, first, of the exclusion of all evidence pertaining to the "B Test," a series of tests performed by NCR on a prototype Century 100 Series computer system beginning in April 1968 and lasting for several months. These tests coincided in part with the Renfro-NCR negotiations. Contending that the test data showed that NCR was experiencing "severe difficulties" with the Century 100, Renfro offered the B Test results as proof of the actual operation of the Renfro computer and of Appellant's awareness of these "difficulties" at the same time that it was being represented to Renfro that the Century 100 was "reliable." The trial judge, however, found that the subject prototype had not been shown "to have resembled in any reasonable manner the Century 100 actually delivered to Renfro," so that the relevance of the B Test data was "highly suspect," Tr. 1509. Accordingly, this evidence was excluded.

Renfro claims that this ruling was erroneous on two principal grounds. First, Appellant asserts that the prototype "clearly" was constructed to the same specifications as the production models and was "virtually identical" to the system installed at Renfro. Second, Appellant argues that, in the context of this transaction, NCR had a duty to disclose any unfavorable test results which might have affected the decision to enter into a lease agreement, because it involved a new product still undergoing development and so complicated that the customer was forced to place unusual reliance on NCR's expertise. Brief at 6-11. Thus, Renfro contends that the B Test results were particularly relevant because they were the only test results available to NCR at the time of the negotiations.

Under both North Carolina and federal law,5 the relevance of experimental evidence depends on whether or not the experiment was performed under conditions "substantially similar" to those of the actual occurrence sought to be proved. Perfecting Service Co. v. Product Development and Sales Co., 259 N.C. 400, 412-413, 131 S.E.2d 9, 19 (1963); Mintz v. Atlantic Coast R.R. Co., 236 N.C. 109, 114-115, 72 S.E.2d 38, 43 (1952); Olin-Mathieson v. Allis-Chalmers, 438 F.2d 833, 837 (6 Cir. 1971); Ramseyer v. General Motors Corporation, 417 F.2d 859, 864 (8 Cir. 1969); Derr v. Safeway Stores, Inc., 404 F.2d 634, 638-639 (10 Cir. 1968). If there is substantial similarity, the differences between the test and the actual occurrence ordinarily are regarded as affecting the weight of the test evidence rather than its admissibility. Ramseyer, supra. On the other hand, the differences between the test and the actual occurrences may be such that the trial judge is justified in concluding either that the evidence is totally lacking in probative value as to any material issue, or that the probative value of the evidence is overborne by the danger that introduction of the evidence will tend to confuse the issues, unnecessarily prolong the trial, or create a likelihood of undue prejudice. In such cases, it is proper to exclude the evidence, even though it may in a sense be termed "relevant."6 The burden of satisfying the Court as to the admissibility of experimental evidence, with these considerations in mind, rests with the proponent. See Ramseyer, supra; Derr, supra, at 638 and n. 7.

There was no error in the finding of the Court that the Appellant had failed to meet this burden. The Final Report on the "B Test" emphasized that "(t) he 615-100 System (tested) greatly deviates from a production system." Report at 1. Numerous deviations between the prototype and production model were listed for every major component tested; in some instances, the dissimilarities were so great that major units could be tested only for interface capability. Id. at 22. Furthermore, the conditions under which the prototype was tested differed drastically from an operational environment; see, e. g., Final Test Report at 10 (tests on processor).7 Plainly, under these circumstances, this Court cannot say that the trial judge was erroneous in concluding that Appellant had failed to show similarity between the test conditions and conditions at Renfro's facility.

The infirmity in Appellant's proffer of the B Test data is not cured by the allegation that NCR may have had a duty to disclose to Renfro the state of development of its new computer. The case offered in support of Appellant's argument that the B Test should have been admitted because NCR had such a "duty" was Strand v. Librascope, Inc., 197 F.Supp. 743 (E.D.Mich.1961). There the defendant, a manufacturer of computer components, was found to have withheld critical information about pre-distribution testing of its new computer components from plaintiff, a purchaser. The question of relevancy is not addressed in Strand; it is obvious, however, that the court found the proof of this testing to be relevant in the circumstances of the case. It appears, although it is not certain, that the component tested was exactly the same as the one supplied to the plaintiff, so that proof of differences between defendant's testing and plaintiff's operation was particularly relevant. Here, both the computer system tested and the test conditions varied, a notable point of distinction. But in any event, relevancy of proof was a determination to be made on the circumstances presented. Whatever obligations may have existed, given hypothetical circumstances, the question is whether or not the actual circumstances of the transaction were properly provable by the evidence offered. The determination of the relevancy of such proof was a matter in the discretion of the district judge. Beaty Shopping Center, Inc. v. Monarch Ins. Co. of Ohio, 315 F.2d 467, 471 (4 Cir. 1963). The record reveals no abuse of that discretion.

Several exhibits proffered by Renfro contained references to performance goals set by NCR for the Century 100 in terms of a failure rate stated as mean time between equipment failures (MTBF). MTBF data are regarded as "an indication of reliability." Tr. 1469-70. This evidence was offered to show that NCR had set a company standard for reliability for the Century 100. The trial judge held that no relationship had been established between the goals and the Renfro transaction, and that the goals were therefore irrelevant. For that reason, he excluded all references to the goals from the evidence.

The significance of the NCR performance goals was developed in a lengthy voir dire examination of William H. Cuthbertson, NCR's quality control manager for the Century 100 system. Cuthbertson made it plain that the goals amounted to no more than a "target" or "design objective." Tr. 1428, 1456, 1458. He intimated, further, that they were set artificially high in an attempt to goad the NCR engineers assigned to the Century 100 to produce an improved product. Tr. 1455. The end apparently sought to be achieved by the use of the performance goals was not customer satisfaction or conformity with any industry standard but was, instead, the maximization of NCR's profit position. Since NCR's form contract provided for maintenance at NCR's expense, it was desired that service calls be kept to an absolute minimum. See, e. g., Tr. 1446, 1455, and 1470. There was no testimony that the "goals" represented an "acceptable minimum" reliability value to the industry,8 to NCR, or to any client; and counsel for Renfro admitted that no representations as to MTBF were made by NCR in the course of the negotiations. Tr. 1439. In light of this, this Court agrees that the goals were properly excluded as irrelevant.9

Among the two to three feet of documentary exhibits10 proffered by Renfro were several items which were excluded on the issue of breach of warranty and were admitted only on the issue of NCR's knowledge of possible defects and intent to deceive. These exhibits included (1) a series of reports on tests conducted by NCR on the Century 100 after the conclusion of the B Test (discussed in Part II of this opinion), referred to collectively as the "C Test"; (2) records of product performance meetings at which NCR personnel reviewed the status and progress of the new computer system, often with reference to statistics gathered on the entire system and on individual sites in the field;11 and (3) separate statistical reports containing performance and reliability data on individual computers and on the system as a whole.

Renfro argues vigorously that these exhibits should have been admitted in their entirety on the issue of breach of warranty. It is contended that the statistics collected in the C Test reports and corporate minutes "show affirmatively that this equipment was anything but reliable by NCR's own definition" and that "the Century 100 system as a whole, in all locations, was not living up to the representations and warranties made by NCR to Renfro." Appellant further argues that the trial court used hearsay considerations as the basis of exclusion and that the hearsay rule had no application in view of a pretrial stipulation in which the parties had agreed to the admissibility of the documents, reserving only relevance and materiality as grounds for objection.

From the way in which these arguments are framed, it appears that Appellant misperceives the scope of inquiry on the breach of warranty issue. The questions in the instant case were what warranty, if any, was operative in this transaction and whether or not the operation of the system installed at Renfro conformed to that warranty. Whether or not other warranties were made to other parties and perhaps breached was not at issue. Thus, the only conceivable relevance these exhibits might have had on this issue was to show the actual operation of the computer system installed at the Renfro site.

Contrary to Appellant's argument, it appears that the C Test was not excluded on hearsay grounds but rather because the trial judge thought the test results would be incompetent to show the actual operation of the Renfro computer. Tr. 1510.

The initial impression of the trial judge, and that perpetuated in Appellant's Brief, was that the C Test was conducted on production Century 100 equipment "of the exact type supplied to the defendant" (Tr. 1510) rather than prototype models as had been the case with the B Test. Later testimony, however, revealed significant differences between the C Test equipment and that supplied to Renfro. Although characterized as a "production model," the test machine incorporated several components which were not representative of the final production configuration, and one major component (the disc unit) was a prototype. Tr. 1517-1518. Moreover, the conditions under which the machinery was tested often differed radically from normal operating conditions12 for, in an effort to locate "weak spots" in the system, the NCR engineers made every effort to create equipment failures. Tr. 1749.

As we have stated, the admissibility of an experiment to prove an actual occurrence ordinarily depends on a foundational showing that the two took place under substantially similar conditions. The ruling of the court in the instant case was couched in somewhat broader terms, Tr. 1510; however, in view of the record, it seems clear that no sufficient showing was made and that the exclusion of the C Test on the breach of warranty issue was proper.

The question of competence is likewise raised for the performance and reliability statistics contained in the NCR corporate minutes and field reports. Appellant argues most earnestly that these statistics would in some way have been probative of a breach of warranty, citing several cases in which the actual operation of one product has been used as evidence of the operation or quality of another product. In each of these cases, however, to the extent that the issue of competence is discussed, it appears that the court was satisfied as to the similarity of the two products. North Carolina courts have for many years adhered to the rule that evidence of "other occurrences" is admissible to show the manner of an actual occurrence, but only if a high standard of similarity between the two has been demonstrated. See Perry v. Kelford Coca-Cola Bottling Co., 196 N.C. 690, 146 S.E. 805 (1929) (the essential physical conditions on the two occurrences must have been identical). Thus, for example, in Tennessee Carolina Transportation, Inc. v. Strick Corp., 283 N.C. 423, 196 S.E.2d 711, 721-722 (1973), evidence of the failure of 9 of 150 trailers was admitted to show defects in the entire order where all of the trailers had been constructed to the same specifications, all had been used by a single customer, and most had failed within a relatively short time span. The cases cited by Appellant, to the extent this issue is raised therein, appear to be of the same genre, and certainly none is contrary. See, e.g., Swift & Co. v. Aydlett, 192 N.C. 330, 135 S.E. 141 (1926) (evidence of performance of fertilizer sold in a prior year admitted to show defect in fertilizer sold to plaintiff where both had been sold by the same agent and had been guaranteed to be of the same chemical composition); Standard Paint and Leadworks v. Spruill, 186 N.C. 68, 118 S.E. 891 (1923) (evidence of other paint, warranted as fireproof, admitted for comparison with the paint at issue where all of the paint had been obtained from one source and where the court was satisfied that the two paints were "exactly similar").

Of course, the North Carolina rule of exclusion ordinarily would not have bound the court; the trial judge was free to admit any relevant evidence so long as the hearing was kept within reasonable bounds. United States v. 25.406 Acres of Land, etc., 172 F.2d 990, 995 (4 Cir. 1949). See, e.g., Thomas v. Hogan, 300 F.2d 355, 362 (4 Cir. 1962) (error to exclude evidence solely because of state rule of exclusion). Nevertheless, absent any showing of similarity between the Renfro computer and its manner of operation, and the various computers for which statistics were gathered and incorporated into the disputed exhibits, the exclusion of this evidence for the purpose of showing a breach of warranty would have been proper. No such showing was made.13

Furthermore, the particular evidence proffered here was voluminous and complex. It contained masses of statistics which might well have been confusing to a layman, and absent any standard for reliability (because the goals were excluded, and properly so), we can only wonder what conceivable use a jury might have made of these exhibits in considering how the Renfro computer actually functioned and whether or not it was "reliable."14 In other words, these exhibits had little if any probative value but rather considerable potential for prolongation of the trial and possible confusion of the issues and could well have been excluded for that reason. See Note 6, supra.

It seems from the record, however, although the matter is not free from doubt, that the principal ground relied on by the trial court in excluding these exhibits on the warranty issue was its judgment that they contained inadmissible hearsay. From a review of the authorities cited and other statements in the transcript, it appears that the Court was particularly concerned with whether or not statements made by NCR employees in the disputed exhibits were admissions by the corporation and, thus, outside the scope of the hearsay rule. See Tr. 1338, 1505, 1510.

Appellant contends that any such ruling was erroneous because the parties had agreed in a stipulation, incorporated into the Final Pretrial Order, that the documentary exhibits listed in that Order (including those at issue) were "genuine and, if relevant and material, may be received in evidence without further identification or proof." Final Pretrial Order at P 12. Because the local rules of court require that all objections reserved as to admissibility of documents listed in the pretrial order be set out with particularity15 (except for objections based on relevance or materiality), it is argued that any objection based on hearsay had been waived and, thus, that the exclusion of evidence on hearsay grounds was improper.

The court was well aware of the local rule and, indeed, had had occasion to construe and apply it earlier in the trial. When Appellee explicitly challenged the admission of Renfro's Exhibits 41 through 49 on the ground that they contained inadmissible hearsay, the court ruled that this ground for objection had been waived.16 Tr. 1243. In view of the emphatic stance taken by the trial judge on these occasions as to the proper application of Rule 22(j)(5), the apparent exclusion of the disputed evidence on the ground that it was inadmissible hearsay is enigmatic.17 Certainly, having decided on a construction of the local rule, and this Court expresses no opinion on the propriety of that construction, the judge was bound to apply the rule with consistency.

If this was error, however, plainly the error was harmless. This evidence was properly excludable as incompetent or remote. Moreover, the other evidence offered as to the actual functioning of the Renfro computer was considerable, if not overwhelming; two logs were submitted giving detailed data on the performance, reliability (or lack thereof), and maintenance of the computer system over the several months during which it was in use. In addition, literally reams of testimony were given, much of it bearing directly on the actual experience at the Renfro site, including the period of testing that preceded the first billing. Given the slight probative value of the minutes and reports and the lack of any standards by which the jury could have determined the significance of the statistics contained therein, the exclusion of this evidence can have had little effect on the jury's decision regarding breach of warranty. Thus the limitation on admissibility based on hearsay, although apparently erroneous, was not so weighty as to require reversal.

In conclusion, it appears from the record that the rulings on evidence assigned as error were correct, with the possible exception of the limitation as to use of the corporate minutes and statistical reports. These we think could, and perhaps should, have been excluded, but not for the reason given by the district judge. In any event, that particular ruling, if error, was harmless. Accordingly, the judgment is affirmed.

AFFIRMED.