Simmons, Inc., a Corporation v. Pinkerton's, Inc., a Corporation and National Surety Corporation, a Corporation , 762 F.2d 591 ( 1985 )

CUDAHY, Circuit Judge.

Defendants-appellants Pinkerton’s, Inc. and National Surety Corporation appeal a judgment against them in an action for property damages arising out of a fire at a warehouse owned by plaintiff-appellee Simmons, Inc. The jury returned a verdict in favor of Simmons and assessed damages in the amount of $971,012.65. We affirm.

I.

Simmons, a manufacturer of residential and commercial bedding, owned a warehouse in Munster, Indiana, which it used as a regional distribution center for its products. In order to protect the warehouse, Simmons entered into a contract with defendant Pinkerton’s, a national company providing investigative and security services, under which Pinkerton’s agreed to provide uniformed guard protection for the warehouse 24 hours a day, 7 days a week. The contract provided that Pinkerton’s would “ensure a professional, reliable and efficient effort to protect its clients’ property and personnel against security hazards.” In addition, Pinkerton’s explicitly accepted liability for all acts of negligence, fraud or dishonesty on the part of its security employees in the performance of their duties, but disclaimed any other liability. Pinkerton’s also represented that its employees were trained both in security and in fire protection.

About July 18, 1978, Pinkerton’s hired William Hayne for the position of security guard. Hayne apparently lied in several instances on his employment application, but the deceptions went undetected since Pinkerton’s failed to check Hayne’s references and other sources of information about him, in contravention of its own policy and procedure manuals. Pinkerton’s also neglected to give Hayne certain types of training, including fire protection training, which its manuals indicated were mandatory for security guards.

On September 25th, 1978, William Hayne reported for work at the Simmons warehouse at 3:00 p.m. The security guard on duty ordinarily makes rounds at designated security points throughout the Simmons warehouse and periodic checks of fire reporting equipment, but because the main door controlling access to a loading dock at the warehouse was broken and fixed in an open position, Hayne had been ordered to remain at a guard desk near the loading dock. At 4:30 p.m., all Simmons’ employees left the warehouse. At about 5:45 p.m., an office cleaning person, Ms. Anna Benedict, reported for work. Hayne and Benedict were the only people present in the loading dock area of the warehouse, and they spent some time talking in an office near the loading dock, where Benedict was working. About 6:00 p.m., Hayne left his station and went to get a drink from some vending machines adjoining a storage area in a different part of the warehouse. He then returned to the loading dock area. A few minutes later, he offered to get a soft drink for Benedict, and then went back to the vending machines to do so. He came back, gave Benedict the drink, and started back to his post. According to Hayne’s testimony, he then discovered a fire in progress. Hayne attempted to put the fire out with several cannister fire extinguishers but was unsuccessful. A reel fire hose was available in the area, but Hayne was not trained to use the hose and was unable to operate it properly. The Munster Fire Department arrived and contained the fire, but not before damage to the building and extensive damage to the bedding occurred.

The Chief of the Munster Fire Department determined that the fire originated on or near a wooden partition that separated stored rows of bedding. Because the point of origin was 12 inches above the floor in an area where there was no apparent source of ignition, and in light of other suspicious circumstances, the Munster Fire Chief requested the Indiana Fire Marshal’s Office to investigate whether Hayne might have accidentally or intentionally started the fire. An investigator from that office concluded that the fire was of incendiary origin and was most likely set by Hayne as an “attention getter.” Pinkerton’s requested that Hayne take a polygraph examination, which was performed in Chicago. The results of this examination were inconclusive. Subsequently, Hayne agreed to take a second polygraph examination in Indiana, but did not do so. Nevertheless, when an investigator later asked Hayne whether he had taken and passed the second test, Hayne falsely stated that he had.

Simmons then brought this diversity action against Pinkerton’s, and against the National Surety Corporation as surety for Pinkerton’s. The complaint alleged that Pinkerton’s was liable under several theories: breach of the contract to provide fire protection and security services, failure to use reasonable care in providing such services — particularly with respect to selecting, training and supervising security personnel — and statutory liability under the Indiana Detective Licensing Law. After trial the jury found in favor of Simmons and awarded it the amount of damages claimed. Pursuant to Simmons’ post-trial motion, the trial court awarded prejudgment interest on the damages. Pinkerton’s and National Surety raise several claims of error relating to the trial court’s instructions to the jury, admission of evidence and award of prejudgment interest. We consider each of these claims separately.

II.

The defendants’ first contention is that the trial court erred in instructing the jury that a violation of the Indiana Detective License Law, Ind.Code § 25-30-1-1 et seq. (1982), constitutes negligence per se, when ■that statute is irrelevant to the issues in this case, and does not provide a standard of care creating a private cause of action for the plaintiff. Despite the objections of all the parties, the trial judge gave the jury Instruction No. 5 on the Indiana Detective License Law. The instruction quoted verbatim seven separate sections of that statute, including provisions relating to the short title of the act, the definitions of “licensee” and “private detective business,” the requirements for a license, the grounds for denying, suspending or revoking a license, the requirement that a licensee obtain a surety bond and a provision entitling a licensee to employ unlicensed persons to assist it. The challenged instruction concluded with a statement that if the jury found from a preponderance of the evidence that the defendant violated “the provisions of the statute, such conduct would constitute negligence on the part of the defendant if done without excuse or justification,” and should result in liability if it was the proximate cause of the damages.

The defendants argue, and we agree, that most if not all of the provisions of the Indiana Detective Law quoted in Instruction No. 5 are not relevant to any issue in this case. The bulk of the quoted provisions relate to requirements that licensees must meet in order to obtain and maintain a mandatory license. (The term “licensee” as used in the statute refers to detective businesses, such as Pinkerton’s, not to their employees.) Yet there was no allegation that Pinkerton’s had failed to qualify for or obtain a license, had or should have had its license revoked, had failed to obtain a surety bond or was similarly delinquent. Thus we do not see how the jury could have concluded that Pinkerton’s had “violated” any of these provisions, or that such a violation proximately caused plaintiff’s damages. There was no good reason for giving this portion of the instruction.

The only provision of the Indiana Detective Law that arguably was applicable to the issues in this case is section 25-30-1-11, which provides:

25-30-1-11. Employees of licensee; record; requisites—

(a) A licensee may employ, to assist him in his business as a private detective, as many unlicensed persons as may be necessary. Such a licensee is civilly responsible for the good conduct of each employee while he is acting on behalf of the licensee.

Simmons contends that this section creates a cause of action for damages caused by the wrongful conduct of security guards while on duty, and hence the instruction that referred to the section was proper. Pinkerton’s, on the other hand, argues that including this section as part of the instruction was reversible error because the provision does not establish a statutory standard of care, the violation of which would constitute negligence.

In diversity cases state law determines the substance of jury instructions, while federal law governs the procedure in formulating the instructions and the manner in which they are given. E.g., Morris v. Getscher, 708 F.2d 1306, 1309 (8th Cir.1983); In re Air Crash Disaster Near Chicago, Ill, 701 F.2d 1189, 1199 (7th Cir.1983), cert. denied, — U.S.-, 104 S.Ct. 204, 78 L.Ed.2d 178 (1983). Thus our first task is to determine whether the instruction given accurately reflected Indiana substantive law. Since section 25-30-1-11 (relating to the good conduct of licensees’ employees) is, for this purpose, the pertinent part of the statute, our initial focus is on the meaning of that provision. Unfortunately, our task is made more difficult by our inability to find any Indiana cases construing it. Indeed, the only relevant case that the parties have cited is Stewart Warner Corp. v. Burns International Security Services, Inc., 353 F.Supp. 1387 (N.D.Ill. 1973), which interprets a similar provision of an Illinois statute.

In Stewart-Warner, as in this case, a security guard was accused of starting a fire on a customer’s premises while he was on duty, and the customer sued the company that employed the guard for the value of the goods destroyed. Two of the counts in that case were based on a theory of negligent hiring, and one other was premised on statutory liability. The statutory section at issue, Ill.Rev.Stat. ch. 38, § 201-10(b)(10), provided that a detective agency employer “shall at all times during such employment be legally responsible for the good conduct in the business” of each person it employed.¹ The defendant there contended that the statute did not create any sort of liability over and above that existing at common law. The court held, however, that the purpose of the section was rather to impose on detective agencies liability for loss arising from intentional acts of their employees. Without the statute, liability for such intentional acts could not be premised on a theory of respondeat superior, but could only be established by proving negligent hiring. The court stated that the “good conduct” provision extended liability “to all wrongful acts of employees of detective agencies, regardless of whether they were negligently hired and whether the acts are negligent or intentional, so long as they were committed while the employee was actually on the job.” 353 F.Supp. at 1389.

Pinkerton’s has attempted to distinguish Stewart-Warner on several grounds, in order to avoid the conclusion that the Indiana provision, which is almost identical to the Illinois statute construed in that case, imposes similarly broad liability. First, defendant contends that in Stewart-Warner the statute was applicable only because the plaintiff had stated a cause of action for negligent hiring, whereas in this case Simmons cannot claim negligent hiring because of a peculiarity of Indiana law that we discuss at some length in Part III. Assuming for the moment that this limitation on Simmons’ claim based on Indiana law is well founded, the suggested distinction still fails. As noted, the opinion in Stewart-Warner makes it clear that the statutory liability for wrongful conduct exists regardless whether the employee was negligently hired. Thus, if Pinkerton’s is correct that Simmons may not predicate liability on wrongful hiring, application of the statute would be clearly meaningful in that it might provide for liability where there otherwise would be none.

Pinkerton’s also argues that the expansive interpretation of the Illinois good conduct provision in Stewart-Warner is inappropriate with respect to the Indiana provision because the Illinois statute includes many sections relating to employee qualifications and procedures that are absent in the Indiana Detective Act; hence presumably the intended focus of the Illinois Act is different from the Indiana statute. We do not find this argument persuasive. Any additional standards governing employees of licensees in the Illinois Act — which indeed could possibly create additional bases for liability not present in the Indiana scheme — do not affect the basic point that the only provision on which statutory liability was predicated in Stewart-Warner was the good conduct provision also at issue here. We see no evidence in Stewart-Warner that the court’s interpretation of the good conduct section was affected by its awareness of the other employee qualification sections or by some construction of the general purpose of the Illinois Act that might not be applicable to the Indiana statute. Rather, the court’s analysis was essentially a plain meaning interpretation of the words of the good conduct provision, one we find fully applicable to the virtually identical section here. The meaning of a statutory rule that an employer is civilly responsible for its employees’ good conduct in this context seems pretty clearly to be that the employer is liable when an employee engages in bad conduct. Except for this meaning, we see no purpose to be served by including such a provision in the statute, given the liability already established under various more restrictive common law theories.

Our view that section 25-30-1-11 of the Indiana Detective License Law imposes liability on detective agencies for the wrongful conduct of their employees does not necessarily lead to the conclusion that the instruction on that section given here was proper. To the contrary, the form of the instruction seems unnecessarily confusing. As we have noted, section 25-30-1-11 is apparently intended to impose liability for an employee’s intentional torts. The effect of the section thus differs from the impact of respondeat superior. And an instruction that merely recited this section (presumably with any appropriate supplemental definitions of terms or explanations) would be appropriate. But the addition of language purporting to make a violation of the section equivalent to negligence makes no sense, since the good conduct provision does not set forth any standard of care. Thus we do not understand how a detective agency can be thought to “violate” a statute establishing its liability for the wrongful (and presumably intentional) acts of its employees; that liability simply exists if an employee’s actions do not constitute “good conduct.” Hence we think that the form the instruction took was erroneous.

We note, however, that the error with respect to the “good conduct” section is subject to examination under Rule 61 of the Federal Rules of Civil Procedure and should be disregarded if it does not affect the substantial rights of the parties. See Alloy International Co. v. Hoover-NSK Bearing Co., 635 F.2d 1222, 1226-28 (7th Cir.1980). An error in giving (or refusing to give) a particular instruction will not be considered reversible error unless, considering all the instructions, the evidence and the arguments that the jury heard, it appears that the jury was misled or did not have a sufficient understanding of the issues and its duty to determine them. See Qasem v. Kozarek, 716 F.2d 1172, 1180 (7th Cir.1983); Alloy International, 635 F.2d at 1226-27. The instructions must be evaluated “as a whole, in a common sense manner, avoiding fastidiousness, inquiring whether the correct message was conveyed to the jury reasonably well.” Reversal is inappropriate unless the jury’s understanding of the issues was seriously affected to the prejudice of the complaining party. Wilk v. American Medical Association, 719 F.2d 207, 218-19 (7th Cir.1983), cert. denied, — U.S.-, 104 S.Ct. 2398-99, 81 L.Ed.2d 355 (1984).

Under these standards,² we do not think that the form of the instruction given here constitutes reversible error. The trial court gave a number of other (correct) instructions as to Pinkerton’s liability for Hayne’s wrongful conduct in negligently or intentionally setting the fire, among other theories, and we think the jury was sure to understand that the additional instruction that Pinkerton’s was civilly liable for the good conduct of Hayne did not impose any liability unless Hayne had acted wrongfully. Indeed, Pinkerton’s does not seriously contend otherwise, but merely argues that the recitation of section 25-30-1-11 was “unnecessary” (as well as confusing in the sense discussed above) since Pinkerton’s “did not deny legal responsibility for any wrongful act that William Hayne had committed at the time of the loss.” (Pinkerton’s Br. at 19.) Thus we do not believe the portion of the instruction relating to the statutory good conduct provision, although erroneous, prejudiced Pinkerton’s.

Our analysis is only slightly different with respect to the other statutory sections quoted in Instruction No. 5, which relate to qualifications of licensee detective agencies, grounds for denial, suspension or revocation of licenses, bonds, and so on. As noted, these provisions appear to be entirely irrelevant to the case. It is conceivable that some of them could describe a statutory duty the violation of which could give rise to liability for negligence, but we are unable to find any indication that any evidence was presented at trial that these provisions were violated. The trial court’s rationale for including them in the instruction apparently is reflected in its statement that the “quoted sections established the applicability of the statute to enterprises engaged in security services and outlined the requirements such entities must satisfy under the Statute.” (District Court Memorandum and Order, March 6, 1984, at 21.) But while a few quoted sections did establish the statute’s applicability to Pinkerton’s, the same end seemingly could have been accomplished with less confusion by a direct statement to that effect by the trial judge, and again, it is improper to outline the “general requirements” a detective agency must meet if there is no contention or evidence in the record that the agency has not done so. Thus the recitation of these provisions as part of Instruction No. 5 was error as well.

We do not think, however, that the error is significant enough to warrant reversal because a common sense examination of the instructions as a whole convinces us that it is quite unlikely that the inclusion of the extraneous material confused or misled the jury to the prejudice of the defendant. Pinkerton’s suggests that the jury may have been confused about the term “licensee” under the statute, incorrectly believing that security guard Hayne, not Pinkerton’s, was required to meet the various licensing requirements. It contends that various evidence relating to Pinkerton’s omission of a security check on Hayne before he was hired, as well as Hayne’s dishonesty in his application for employment, might have been taken by the jury to be “violations” of the (inapplicable) statutory provisions, and thus to form the basis for the belief that Pinkerton’s was negligent.

This hypothesis depends on a number of doubtful assumptions: 1) that the jury ignored the clear language of the statute which distinguishes between a “private detective business” that must obtain a license and thus becomes a “licensee,” and “employees of licensees,” who may be unlicensed persons, and so erroneously applied the licensee requirements to security guard Hayne; 2) that the jury then stretched the clearly inapplicable provisions to fit evidence presented on different matters — for example, that it confused the statutory requirement that Pinkerton’s apply for a license (about which there was no evidence) with Hayne’s application for employment; and 3) that the jury then concluded that Hayne’s “failure” to meet some or all of the inapplicable licensing requirements proximately caused the damage to Simmons’ warehouse. (Pinkerton’s argument as to prejudice also assumes, of course, that the jury would not have found Pinkerton’s liable on one of the other theories— that Hayne intentionally or negligently set the fire, or was negligently hired, trained or supervised so that he was deficient in his attempts to extinguish it — but for its confusion stemming from the erroneous instructions on statutory liability.³) We find this chain of reasoning quite implausible, and therefore find the erroneous instruction — albeit incomprehensible — to be harmless error in the present context.

III.

Pinkerton’s next complains of the introduction of various evidence that related to industry-wide, as well as to Pinkerton’s own, practices with regard to hiring, training and supervising security guards, and Pinkerton’s deviation from these practices in hiring, training and supervising William Hayne. At trial, Pinkerton’s objected to this evidence on the basis that it was not relevant under Federal Rule of Evidence 402, and that its probative value was substantially outweighed by its prejudicial effect and thus should have been excluded under Federal Rule of Evidence 403. The core of its argument is the proposition that all of Simmons’ theories of recovery — tort, contract and statutory — have in common the necessity of proof that Hayne acted wrongfully at the time of the fire at Simmons’ plant. Because Pinkerton’s has stipulated that Hayne was acting within the scope of his employment at the time of the loss, it argues that Simmons need only have proven that Hayne’s wrongful act at that time proximately caused the damage. This would have carried Pinkerton’s burden under the tort principle of respondeat superior. Thus, any additional evidence potentially relevant to other tort, contract or statutory theories, such as Hayne’s background and the circumstances of his hiring, training and supervision, was unnecessary to Simmons’ success on the merits, and could only serve to prejudice the jury against Pinkerton's.

In support of this argument, Pinkerton’s cites two cases which it contends establish that Indiana law prohibits evidence that an employer negligently hired, trained or supervised an employee, where the employer stipulates that the employee’s allegedly wrongful act was within the scope of employment. In the first, Lange v. B & P Motor Express, Inc., 257 F.Supp. 319 (N.D.Ind.1966), the district court sitting in diversity refused to allow plaintiffs to amend their complaint to assert a new cause of action against an employer for negligent hiring, where the original complaint sought to recover from the employer for personal injuries caused by its employee’s negligent driving under traditional principles of respondeat superior. The employer in that case had admitted that the employee was operating within the scope of his employment at the time of the collision that prompted the suit. The proposed amendment to the complaint would have alleged direct negligence on the part of the employer by reason of its hiring or retaining in employment a person whom it knew or should have known was a negligent and unsafe driver.

The court first reviewed the general state of the law on the negligent hiring theory, concluding that the theory had only been applied in “special situations” in which traditional agency principles would probably not allow liability. Where a respondeat superior theory was applicable, a negligent hiring theory ordinarily would not be. A respondeat superior case requires proof of a wrongful act or omission by the employee causing the plaintiff’s injuries. Such proof generally would be sufficient to establish the employer’s liability under respondeat superior, regardless whether the employer used due care in hiring the employee. Thus the additional evidence necessary to prevail on the negligent hiring count, indicating that the employer had reason to know of the employee’s dangerous propensities — for example that the employee previously had driven negligently — would not only be superfluous but would be prejudicial on the issue whether the employee actually drove negligently at the time of the plaintiff’s loss. On the other hand, in situations in which a respondeat superior theory would not be successful, despite the employee’s wrongful act — as, for example, where the bad act was intentional and not considered to be within the scope of employment — then proof that the employer should have known of the employee’s dangerous traits and so should not have hired her “takes on independent significance.” 257 F.Supp. at 321-23.

Following this reasoning, the court in Lange refused to permit the plaintiff to attempt to prove negligent hiring in addition to its original claim under respondeat superior, despite the court’s recognition that Indiana apparently does allow a negligent hiring cause of action. Since the employer in Lange admitted scope of employment, proof of the employee’s past negligence would only “reinject the agency issue under the guise of negligence in hiring which adds to the burden of proof while prejudicing the defendant’s case on the issue of the driver’s negligence at the time of the accident.” Id. at 324.

A similar result was reached by the court in Tindall v. Enderle, 162 Ind.App. 524, 320 N.E.2d 764 (3d Dist.1974). In that case an employee of the defendant tavern-owner shot two patrons of the tavern, killing one of them. In the plaintiffs’ subsequent action for wrongful death and negligence against the tavern and the employee, the tavern stipulated that the employee had fired the shots while acting within the scope of his employment. The plaintiffs wanted to introduce evidence that the employee had previously assaulted others, including patrons of the tavern, to show that the tavern was negligent in hiring and retaining the employee despite knowledge of his violent propensities; the tavern filed a motion in limine to exclude this evidence. The court granted that motion, and the plaintiffs appealed. Following the reasoning in Lange, supra, the court held that the evidence had been properly excluded. The court observed that the import of the “ ‘separate’ cause of action [for negligent hiring] generally arises only when an agent, servant or employee steps beyond the recognized scope of his employment to commit a tortious injury upon a third party.” 162 Ind.App. at 529, 320 N.E.2d at 767-68. By contrast,

that theory is of no value where an employer has stipulated that his employee was within the scope of his employment. The doctrine of respondeat superior provides the proper vehicle for a direct action aimed at recovering the damages resulting from a specific act of negligence committed by an employee within the scope of his employment.

Id. at 530, 320 N.E.2d at 768 (footnote omitted). See also Shipley v. City of South Bend, 175 Ind.App. 459, 372 N.E.2d 490 (3d Dist.1978).

We first consider the application of these cases to the evidence that Pinkerton’s failed to adequately train and supervise security guard Hayne. Both Lange and Tindall arguably can be distinguished simply on the basis that neither of those cases involved a claim of improper training or supervision, but only negligent hiring or retention based on the employee’s dangerous propensities. But even if the principle of Lange and Tindall fully applies, we believe that evidence of improper training and supervision was properly admitted in this case.

Pinkerton’s argument is that, as in Lange and Tindall, proof that Hayne acted wrongfully at the time of the fire is a common element to all of Simmons’ theories, and so additional evidence of improper training and supervision should have been excluded as both unnecessary and prejudicial. We do not agree with the premise of this argument. One of Simmons’ theories of recovery was based on its contract with Pinkerton’s, in which Pinkerton’s promised it would “ensure a professional, reliable and efficient effort to protect its clients’ property and personnel against security hazards.” There was abundant evidence at trial that both industry standards and Pinkerton’s own training manuals clearly considered fire prevention and fire fighting to be one of a security guard’s most important duties, and so provided for certain (fairly elaborate) training procedures in these matters. Despite these standards, Pinkerton’s gave William Hayne virtually no effective fire-fighting training. Similar evidence was presented with respect to supervision. We think it self-evident that these failures would be inconsistent with a “professional, reliable and efficient effort” to protect Simmons’ property, and hence could be considered breaches of the contract.

More important for immediate purposes, proof that such breaches of contract proximately caused Simmons’ damages would not necessarily involve proof that William Hayne acted wrongfully at the time of the fire, thus unnecessarily duplicating the evidence necessary to show liability under respondeat superior. For example, the jury might well have disbelieved Simmons’ evidence that Hayne started the fire, either intentionally or negligently. Further, the jury was free to believe that Hayne was not negligent in failing to discover or adequately fight the fire. Under these circumstances, the jury would have to find in favor of defendant Pinkerton’s on the tort and statutory theories, since Hayne would not have committed any wrongful act. This would not mean, however, that the jury could not find liability on the contract. Hayne might have done everything in his power to combat the fire, acting reasonably and adequately — even heroically — and yet Pinkerton’s utter failure to train him properly, in breach of its contract, might have proximately caused Hayne’s inability to fight the fire effectively, and the resulting damage to the warehouse.

Indeed, this is not mere theoretical speculation: Hayne testified at trial that after discovering the fire and attempting to fight it with hand-held fire extinguishers, he then attempted to use a reel-type fire hose, but could not get it to operate. Simmons presented evidence that the reel fire hose was fully operational (and would have been far more effective in putting out the fire than the fire extinguishers). Hayne, however, had not been trained to use it and so did not realize that the hose had to be completely unreeled before it would operate. Thus, the jury might properly have concluded that Pinkerton’s breached its contract with Simmons by failing to train Hayne and so provide a “professional, reliable and efficient effort,” even though Pinkerton’s was not liable under respondeat superior principles. Therefore, the evidence of improper training and supervision was not unnecessarily duplicitous and should not have been excluded under the reasoning of Lange and Tindall.

The evidence as to negligent hiring presents a more difficult question because this was the sort of evidence directly at issue in Lange and Tindall. The district court found those decisions distinguishable on the basis that this case not only involves allegations of liability based on simple negligence/respondeat superior, but also involves allegations of liability premised on theories of breach of contract, statutory liability and intentional tort, to which the negligent hiring evidence was relevant. As a result, the court allowed evidence pertaining to industry-wide and to Pinkerton’s standard hiring practices, and to the (allegedly deficient) hiring procedure used with respect to Hayne, which failed to uncover information about Hayne’s background that presumably made him unsuitable for employment as a security guard.

We are not sure that the additional theories of liability involved in this case alone present a sufficient basis for distinguishing Lange and Tindall. Unlike the evidence of improper training and supervision, evidence of negligent hiring apparently only adds to plaintiffs burden of proof without accomplishing any independent purpose. Assuming that Hayne was indeed dishonest, unreliable, and so on, and that Pinkerton’s breached its contract and/or was negligent in failing to discover these facts at the time it hired him, apparently such a failure could not be the proximate cause of Simmons’ damages unless Hayne acted wrongfully at the time of the fire. Therefore, proof of Hayne’s wrongful conduct would be necessary for both the respondeat superior theory and the negligent hiring theory; if wrongful conduct were established, it would ensure success on the respondeat superior theory, thereby making additional evidence of negligent hiring redundant, and unnecessarily prejudicial. Thus, such evidence may fall within the spirit, if not the letter, of the Lange-Tindall rule.⁴

Nevertheless, we do not think the rule is so inflexible as to require reversal in this case. First of all, we do not believe Lange and Tindall stand for the proposition that whenever proof of a given proposition is sufficient to establish liability under one theory, no additional proof necessary to succeed under another theory may be admitted. Rather, we think a common-sense assessment must be made in each such case to determine the purpose for which the “additional” evidence is sought to be admitted, its probative value and its potential prejudicial impact. In Lange and Tindall, it was obvious that no purpose could be served by allowing evidence of the employee’s prior bad acts except to prejudice the jury and so it was sensible to exclude it. Yet the courts in both those cases acknowledged that the result would be different were the evidence relevant to some legitimate consideration so that it took on some independent significance (for example, were the plaintiff to allege an intentional tort, Lange, 257 F.Supp. at 322-23; Tindall, 162 Ind.App. at 529, 320 N.E.2d at 767-78, or pursue punitive damages, Tindall, 162 Ind.App. at 530, 320 N.E.2d at 768.) Here, although evidence as to negligent hiring may have been technically unnecessary given proof of wrongful conduct in setting or fighting the fire, it was highly relevant to several alternative but intertwined contract and tort theories. (Presumably, for example, it established Pinkerton’s breach of its promise to ensure a “professional effort” to protect Simmons’ property.) In other words, we think the evidence has sufficient “independent significance” that plaintiff was entitled to proceed on its alternative theories despite the overlap in evidence.⁵

Moreover, we think the potential prejudicial impact of the evidence here was relatively slight. Not only does this factor weigh in favor of admission but it also persuades us that, even if the evidence was erroneously admitted, no reversible error occurred. The evidence as to Hayne’s background consisted of the information that he had lied on his application to Pinkerton’s, presumably to hide such negative information as that he had not attended high school, had been discharged from the Army after going AWOL several times, had been fired from several jobs for various reasons (including, oddly enough, from Pinkerton’s, where he had worked as a guard previously but was discharged for missing work) and had spent time in jail— on one occasion for an incident involving driving with liquor in a car, and on another for writing bad checks. While this information certainly reflected unfavorably on Hayne (and presumably on Pinkerton’s), it is hardly so shocking or offensive as to elicit a strong emotional reaction from the jury. Indeed, Hayne explained these circumstances at some length, for example, attributing much of his instability in the Army and his employment history, as well as the incident involving writing checks on insufficient funds, to the fact that his mother had a permanent illness requiring him to contribute much of his time and financial resources to caring for her and his sisters.

Most important, there was no evidence that Hayne had.ever been involved with any other fire or arson investigation, and therefore, quite unlike the situation presented in Lange and Tindall, the jury had no similar past wrongful conduct from which it might improperly infer that Hayne had a propensity for setting fires. (To the contrary, it seems the information that Hayne had previously worked for Pinkerton’s for almost a year without incident might well have cast doubt on Simmons’ theory that Hayne had set the fire as an “attention-getter.”) For these reasons we hold that the admission of evidence relating to hiring practices and Hayne’s background was not reversible error under the circumstances presented here.

IV.

Pinkerton’s next argues that the district court erred in admitting Hayne’s testimony that he falsely told a Pinkerton’s investigator that he had taken and passed a polygraph examination regarding the circumstances of the fire. After the fire, Pinkerton’s asked Hayne to take a lie detector test in Chicago. He took the test, but apparently it was inconclusive for extraneous reasons. Hayne then agreed to take a second test in Indiana, but failed to make arrangements to do so. Nevertheless, Hayne later told a Pinkerton’s investigator that he had taken the second test in Indiana and had passed it. At trial, over Pinkerton’s objection, Simmons’ attorney elicited from Hayne his admission that he had lied about this matter. Unfortunately this testimony was sufficiently ambiguous that it could have implied either that Hayne had lied about taking the test, or that he had lied about passing it.⁶

Although “the rule in the Seventh Circuit is clear” that admission or exclusion of polygraph evidence is within the sound discretion of the trial judge, United States v. Rumell, 642 F.2d 213, 215 (7th Cir.1981), for various reasons this discretion has most often been exercised in favor of exclusion, in this circuit, see id. at 215, as well as in others, deVries v. St. Paul Fire and Marine Ins. Co., 716 F.2d 939, 945 n. 8 (1st Cir.1983). Several circuits apparently exclude polygraph results entirely, id. Indeed, because of the perceived danger of prejudice, some courts also exclude evidence of a person’s willingness or unwillingness to take a polygraph examination, id. at 944-45. For these reasons, Pinkerton’s argues, it would have been error to admit the results of any polygraph that Hayne took as substantive evidence of the cause of the fire at the warehouse. The trial court did not admit Hayne’s testimony as substantive evidence, however, but rather admitted it under Federal Rule of Evidence 608(b) to impeach Hayne’s credibility through cross-examination about a specific instance of his conduct (his lie about the polygraph test) bearing on his character for truthfulness or untruthfulness. Pinkerton’s contends that this was an improper basis for admission, because “a witness may not be impeached by contradiction as to collateral or irrelevant matters elicited on cross-examination,” United States v. Lambert, 463 F.2d 552, 557 (7th Cir.1972). Since the test for whether a matter is collateral is “whether the party seeking to introduce it for purposes of contradiction would be entitled to prove it as part of his case,” id., and since Simmons presumably could not introduce evidence of the polygraph examination as a substantive part of its case, Pinkerton’s concludes that the “collateral evidence rule” precludes Simmons from impeaching Hayne in any manner that included a reference to whether he took a polygraph test, or whether he passed such a test.

We disagree. Assuming, arguendo, that testimony relating to the polygraph exam was inadmissible as substantive evidence, Pinkerton’s has failed to note distinctions between this case and the authority on which it relies which seem to make the collateral evidence rule completely inapplicable here. The rule apparently was developed in conjunction with a particular type of impeachment — impeachment by contradiction. Impeachment by contradiction simply involves presenting evidence that part or all of a witness’ testimony is incorrect. Thus if an eyewitness to an auto accident testifies that the car that caused the accident was red, impeachment by contradiction relies on evidence that the car actually was yellow. The inference to be drawn is not that the witness was lying, but that the witness made a mistake of fact, and so perhaps her testimony may contain other errors and should be discounted accordingly.

Of course, a particular misstatement may or may not be probative of the witness’ general accuracy, depending on the circumstances, and thus may or may not be worth the time it takes to establish it. For this reason the collateral evidence rule developed. In the above example, assuming the color of the car was not directly relevant to any substantive issue in the case (e.g., if the identity of the car were stipulated), it presumably would not be worth the fact finder’s time to entertain a “mini-trial” on the issue of the car’s color, simply to prove that the witness was mistaken as to this fact. Thus, while the accuracy of a witness’ perception or memory can always be tested through traditional cross-examination techniques, the collateral evidence rule limits the extent to which the witness’ testimony about non-essential matters may be contradicted by extrinsic proof. In short, if a matter is collateral (that is, if it could not be introduced into evidence as substantive proof) then it cannot be proven simply to contradict the witness’ testimony for impeachment purposes. 3 J. Weinstein and M. Berger, Weinstein’s Evidence II 607[05], 607-61 — 607-72 (1984).

But the collateral evidence rule does not, as Pinkerton’s contends, limit the scope of all types of impeachment by cross-examination to only those matters that could be proven as a substantive part of a case; rather it merely precludes extrinsic evidence of certain facts that would impeach by contradiction. Various types of impeachment are allowed, despite the recognition that the subject-matter of the impeachment could not be used substantively, precisely because a witness’ credibility is always an important consideration. Impeachment by prior inconsistent statement is one example; another is the type of impeachment at issue here — an attack on a witness’ character for truthfulness through cross-examination as to specific instances of the witness’ previous conduct (which, of course, have not been the subject of his direct examination).

Federal Rule of Evidence 608(b), which governs this type of impeachment, is exactly tailored to strike the balance that the collateral evidence rule was designed to achieve with respect to impeachment by contradiction. Rule 608(b) allows cross-examination of a witness about specific instances of her past conduct, if probative of truthfulness or untruthfulness, but prohibits the proof of such conduct by extrinsic evidence. Thus, as with the collateral evidence rule, a relevant fact bearing on the witness’ credibility — in this case, that he has lied in the past or acted in some other manner that casts doubt on veracity — may, in the discretion of the trial judge, be considered sufficiently important and probative to be elicited on cross-examination; yet, because of the dangers of confusion, prejudice, waste of time; and so on, that would be inherent if a “mini-trial” on the existence of that fact were allowed, extrinsic evidence of the matter is prohibited.⁷

Applying these principles to the testimony at issue here, it becomes obvious that the collateral evidence rule — in one sense now incorporated into Rule 608(b)— does not prohibit it. The testimony concerned an incident in which Hayne lied about having taken a polygraph examination. Whether or not Hayne took such a test or whether or not he passed may not have been admissible as substantive evidence. Nevertheless, in the discretion of the court, the “specific instance of conduct” in which he lied was, understandably, considered probative of his character for truthfulness or untruthfulness. Therefore, it was proper to allow Simmons to inquire about the incident while cross-examining Hayne. Of course, had Hayne denied the lie, the bar on the use of extrinsic evidence of specific instances of conduct to attack credibility in Rule 608(b) would have prohibited Simmons from proving the lie through the testimony of other witnesses. This sort of proof, however, was not offered. Clearly, the testimony carried some danger of prejudice, and just as clearly the trial court was obligated to balance this risk against the probative value of the testimony. The court did so, and we think it properly exercised its discretion in deciding that the testimony could be elicited under Rule 608(b).⁸

V.

Pinkerton’s remaining contentions, relating to damages and interest, do not merit extended discussion. The trial court instructed the jury that the measure of damages for the partial destruction of personal property is the difference between the fair market value of the property immediately before, and immediately after, the destruction. Damages for the complete destruction of personal property, on the other hand, are measured by the reasonable (fair market) value of the property at the time of its destruction. Pinkerton’s acknowledges this to be a correct statement of Indiana law. See, e.g., Bottoms v. B & M Coal Corp., Ind.App., 405 N.E.2d 82, 93 (4th Dist.1980).

Nevertheless, Pinkerton’s contends that the trial court should have adopted a different measure of damages in this case, because Simmons is a manufacturer, which was able to replace the destroyed goods out of its existing stock, and which claimed no lost sales or profits. Under these circumstances, says Pinkerton’s, “fair market value” — in this case, Simmons’ selling price — is an inappropriate measure because the selling price includes a profit margin, which was not yet “earned” at the time the merchandise was sitting in the warehouse. Thus Pinkerton’s would have us award Simmons not the selling price but only a cost of remanufacturing the goods (which in Pinkerton’s view would not include “profit”).⁹ According to the Restatement of Torts, the general rule as to manufacturers’ profits and its rationale are as follows:

From the time when a chattel is manufactured to the time of its actual use, there may be many markets in which it is sold. Thus, different prices are paid by the wholesaler, the retail dealer and the consumer. Since the measure of recovery is determined by the harm done, the market that determines the measure of recovery by a person whose goods have been taken, destroyed or detained is that to which he would have to resort in order to replace the subject matter. Thus the consumer can recover the retail price; the retail dealer, the wholesale price. The manufacturer, who does not buy in a market, receives his selling price.

Restatement (Second) of Torts, § 911, comment (d) (emphasis added). See also H.K. Porter Co. v. Halperin, 297 F.2d 442 (7th Cir.1961) (Illinois law). Pinkerton’s admits it can cite no Indiana law to buttress its imaginative argument that we should discard the fair market value — selling price measure in this case. None of the other authority it offers is persuasive since it involves materially different questions, such as common carriers’ liability to shippers, or wholesalers’ damages. Thus we perceive no reason based on authority to hold that the court’s instruction on damages was erroneous. Simmons deducted from the list price the cost of selling the inventory, thus taking account of the fact that the inventory was still at the manufacturer’s level in the chain of distribution, not at the wholesaler’s. The manufacturer’s “profit margin,” on the other hand, seems properly attributable to the manufacturing process, not primarily to the process of distribution. The fact that Simmons can replace the goods by manufacturing should not affect a profit margin attributable to the manufacturing process. The profit margin is part of the value as inventory, even though, like the rest of the price, it is not realized until sale.

Pinkerton’s also complains about the lower court’s amendment of the judgment to include an award of prejudgment interest from the date of the fire to the date of the entry of judgment. 28 U.S.C. § 1961 (1982) provides for an award of interest on a money judgment whenever the law of the state in which the court sits permits such an award. Although the statute explicitly refers only to post-judgment interest, this reference is not intended to affect plaintiff’s entitlement to prejudgment interest under state law. In re Air Crash Disaster Near Chicago, Illinois, 480 F.Supp. 1280, 1282 (N.D.Ill.1979), aff'd, 644 F.2d 633 (7th Cir.1981). Prejudgment interest pursuant to Indiana law is allowable on judgments in diversity cases. Price v. Amoco Oil Co., 524 F.Supp. 364 (S.D.Ind.1981).

Under Indiana law, prejudgment interest is proper when damages are ascertainable in accordance with fixed rules of evidence and accepted standards of valuation at the time damages accrue, Rauser v. LTV Electrosystems, Inc., 437 F.2d 800 (7th Cir.1981); Indiana Industries, Inc. v. Wedge Products, Inc., Ind.App. 430 N.E.2d 419, 427 (3d Dist.1982), but is not appropriate when the jury must use its best judgment to assess the amount for past and future injury or elements not measurable by fixed standards of value. N.Y., Chicago & St.L.Ry. Co. v. Roper, 176 Ind. 497, 96 N.E. 468 (1911). Pinkerton’s argues that under this test Simmons is not entitled to prejudgment interest, since here there was a bona fide dispute over the amount of damages that should have been awarded for the loss of Simmons’ inventory.

The trial judge concluded that “the damages as presented and adopted by the jury in its verdict were a function of mathematical computation involving the subtraction of salvage proceeds from actual and average sales prices for the various inventory items.” Dist.Ct.Op. at 35. We do not think the fact that Pinkerton’s disputed the use of various sales lists at trial, for example because some price lists were from after the time of the fire, invalidates the claim for prejudgment interest. We do not agree with Pinkerton’s characterization of Indiana’s “fixed and ascertainable” standard as meaning that whenever more than one figure represnting damages — in this case the value of inventory — could have been adopted, no prejudgment interest may be awarded. Pinkerton’s essential contention that the principal amount of damages must be stipulated, or determinable without trial, is amply refuted by the eases construing Indiana law. In many cases prejudgment interest has been allowed even though the fact finder had to use some degree of judgment in measuring damages. See, e.g., Luksus v. United Pacific Ins. Co., 452 F.2d 207 (7th Cir.1971) (interest awarded from various dates that sums owed for labor, services, rental of equipment, bonuses, etc. were later found to be due); N.Y, Chicago & St.L.Ry. Co. v. Roper, 176 Ind. 497, 96 N.E. 468 (1911) (interest award upheld where measure of damages after fire was fair market value of house); Indiana Industries, Inc. v. Wedge Products, Inc., Ind.App., 430 N.E.2d 419 (3d Dist.1982) (interest awarded for inventory even though original claim substantially differed from damages subsequently sought, and from those awarded); N.Y. Central Ry. Co. v. Churchill, 140 Ind.App. 426, 218 N.E.2d 372 (2d Div.1966) (interest allowed where fair market value of tractor, though disputed, was easily ascertainable with a degree of certainty); Kuhn v. Powell, 61 Ind.App. 131, 111 N.E. 639 (1916) (interest allowed despite dispute over amount due per bushel of corn under unwritten contract). Thus, we interpret Indiana’s standard to mean only that damages must be ascertainable as of a particular time (not actually ascertained prior to trial) according to known standards of value (not liquidated amounts). See Courtesy Enterprises, Inc. v. Richards Labs., Ind.App., 457 N.E.2d 572 (3d Dist.1983). Of course, assessment of the degree to which the jury must use its judgment in ascertaining damages may in close cases be difficult, but we cannot say the district court abused its discretion in deciding that prejudgment interest was appropriate here.

Finally, defendants complain about the trial court’s refusal to direct a verdict for the National Surety Corporation for any damages in excess of the amount of its $5,000 surety bond in favor of Pinkerton’s. Since surety liability is contractual and cannot exceed the amount of liability expressly assumed in the surety contract, Orange Co. v. Brown, 181 Ind.App. 536, 393 N.E.2d 192 (1979), the lower court saw no need to direct a verdict to achieve that result. Since no party on appeal contends that National Surety’s liability should be greater than $5,000, neither do we.

For the foregoing reasons the judgment is Affirmed.

1 . The Detective and Investigators Act that contained the "good conduct" section subsequently was incorporated as §§ 2601-2639 & 2701 of ch. Ill of Ill.Ann.Stat., and then repealed by P.A. 83-1069, § 33, effective Jan. 5, 1984. The new Illinois Private Detective and Private Security Act is Ill.Ann.Stat. ch. Ill, §§ 2651-2680 & 2701 (Smith-Hurd Supp.1984-85).

2 . Although the dissent makes an interesting argument against applying the federal harmless error standard in this federal diversity case, it appears that the law is otherwise. We believe the better view is that summarized by Professors Wright and Miller:

It has been suggested that in diversity cases a federal court should apply the standard of harmless error followed in the state courts [citing R. Traynor, The Riddle of Harmless Error, at 47-48 (1970) ]. This seems both unrealistic and unsound. It is unrealistic because of the subjective nature of the problem and the abstract way in which rules about harmless error necessarily are formulated. There is no advantage in substituting the abstractions favored in the state court for those that the federal courts have developed for themselves. It is unsound because the harmless error doctrine, explicitly stated both in Rule 61 and in an Act of Congress, is an important principle of judicial administration that goes to the proper relation between the trial judge and the jury and the proper relation between appellate courts and trial courts. On matters of this kind the federal court is not required to defer to state rules. Suppose, for example, that some state were to follow a rule that a new trial is required for any error whatever, no matter how trivial, technical, and nonprejudicial it might be. A state is free to make its courts citadels of technicality if it chooses to do so, but it cannot impose that choice on the federal courts, "an independent system for administering justice to litigants who properly invoke its jurisdiction.”

11 C. Wright & A. Miller, Federal Practice And Procedure, § 2883, at 279 (1973) (footnotes omitted). We are not persuaded that the result here would be any different if the state harmless error standard were followed, since, among other things, the same rationale that supports our conclusion that any errors here were harmless under the federal standard presumably also means that such errors should not be considered “radically erroneous” under Indiana law. But even if the federal standard is somewhat more lenient, we believe it is properly applied here for the reasons noted.

3 . Of course, where instructions are such that a jury could have based its decision on one of two or more theories, one of which is valid and the others invalid, we cannot speculate about which theory the jury chose. However, in this case we are simply not persuaded by defendants’ arguments that the erroneous instruction could have led to such profound confusion that the jury really might have based its decision on some theoretical, inapplicable breach of the Indiana Detective Law. We think we are entitled to have somewhat higher expectations of the jury.

4 . We note, however, that it is possible to con- . struct a theory that Pinkerton's negligent hiring of security guard Hayne proximately caused Simmons' damages, even though Hayne did not act wrongfully at the time of the fire. Hayne might well have acted as would a reasonable person in fighting the fire, and thus would not have been negligent; Pinkerton’s could not then be liable under the tort theories. Yet it can be argued that Pinkerton's promised more than the furnishing of a reasonably careful person, or even a reasonably careful security guard. Presumably every local security agency promises it will provide a reasonably careful security guard, but here Simmons hired (and paid for) a national company with an established reputation and record. That company made specific representations as to the quality of its guards (for example, one advertisement bragged: "At the heart of Pinkerton's Security Service stands [sic] our guards. All carefully investigated before employment____ Most of them are career employees.” Plaintiff’s Exhibit No. 14.) So perhaps the "quality” of guard promised by Pinkerton’s is somewhat higher than the minimum quality necessary to satisfy the "reasonable security guard” standard of care. And perhaps a more educated guard (Hayne did not attend high school) would have figured out how to use the reel fire hose, or a more responsible, motivated guard (Hayne’s employment and Army records were less than impressive) would not have left his post to chat with another employee and get a soft drink. Thus, Hayne's actions might not be deemed wrongful, and yet one could still be left with a feeling that they contributed to the damage. If so, Pinkerton's breach of contract in hiring William Hayne, when it would not have done so had it followed its own established screening procedures, could be seen as a proximate cause of the damage. Under this reasoning the evidence as to Hayne's background and Pinkerton's hiring practices would not be deemed unnecessary to establish its liability. We do not rest our decision on this theory, since it was not well-developed at trial, but it does tend to lend weight to the conclusion that the Lange-Tindall rationale should not be extended to cover this case.

5 . The dissent flatly asserts that under Indiana law, once Pinkerton’s stipulated that Hayne was acting within the scope of his employment the contract theory was "no longer viable," and thus the evidence relating to Pinkerton’s hiring practices and Hayne’s background was "irrelevant” and "highly prejudicial.” We are unable to discern the rationale for this conclusion. Neither Lange nor Tindall involved a claim of breach of contract, and so neither case can stand for the proposition that an otherwise valid cause of action based on breach of contract completely evaporates when a party also has a legitimate tort claim. That proposition is patently incorrect where the party has contracted for services as the result of express promises about the quality of those services. At the very least, the existence of the contract here is enough to severely undermine Lange and Tindall as persuasive support for declaring a contract theory "no longer viable.”

6 . The testimony was as follows:

Q: And Pinkerton’s asked you to take a lie detector test, didn't they?

A: Yes they did.

Q: And you took that test at their request?

A: Yes I did.

Q: And then later they asked you to take another lie detector test, didn’t they? A: They only asked me to take one, sir.

Q: Do you recall that you told Mr. Robinson after the first lie detector test that you had passed the lie detector test in Indiana?

A: Yes I do.

Q: You told him that story, didn’t you?

A: Yes I did.

Q: Was that true, Mr. Hayne?

A: No, it was not.

Tr. Yol. II, p. 186.

7 . Indeed, even under the collateral evidence test, impeachment by contradiction of a witness’ testimony generally would not be considered collateral — and thus would be allowed — where the impeaching matter indicates the witness’ character for untruthfulness. See 3 J. Weinstein & M. Berger, Weinstein's Evidence 607[05], 607-65 (1984).

8 . In response to Pinkerton’s complaint that the testimony may have unfairly implied that Hayne failed the second polygraph, instead of merely failing to take it, we note that Pinkerton’s shares the blame for this confusion. Once the damaging testimony was elicited, it would have been a simple matter for Pinkerton's to clarify the circumstances of the misrepresentation on redirect. Presumably as a matter of trial strategy, it chose not to do so; thus a complaint about the confusion on appeal is not well-taken.

9 . Pinkerton’s also complains about Simmons’ use of price lists dated in late 1978 and 1979 as evidence of the ordinary selling price of some of the damaged or destroyed inventory. Although Pinkerton’s has not elaborated on this objection, we take its argument to be that since the prices for some of the items on the late 1978 and 1979 lists included as much as a 5% increase over the prices that were in effect in September 1978, when the fire occurred, the later price lists do not accurately reflect the fair market value of those items at the time of the fire. There is some testimony to support the inference that these prices may have been an appropriate measure of fair market value, since the inventory destroyed or damaged in the fire would not have all been sold immediately, but at least some of it might have been sold through 1979, at the new prices. At the very least, the later price lists were sufficiently probative of the value of the inventory to be admissible; so Pinkerton’s argument really is relevant only to the weight the evidence should have been accorded. Pinkerton’s presented this objection to the price lists to the jury through cross-examination of Simmons’ expert on damages, and during closing argument even argued that a specific figure should be subtracted from the damages to account for this problem. Since the evidence based on the price lists was competent, the jury was free to accept it, despite Pinkerton’s arguments to the contrary.