Bonilla v. Volvo Car Corp.

150 F.3d 62, 1998 WL 412991

Opinion

BOUDIN, Circuit Judge.

In these seven related appeals, we are asked to examine numerous rulings and decisions in a complex civil suit brought under the Racketeering Influenced and Corrupt Organizations Act, 18 U.S.C. § 1962 (“RICO”). This opinion addresses the first of the appeals, the principal appeal (No. 97-1135) of Volvo Car Corporation (“Volvo”), and two companion opinions resolve the remaining appeals. 1 We briefly set forth the procedural history of the case as background for all three opinions.

The plaintiffs are a class of individuals and corporations who, between 1983 and 1993, among them purchased 9,077 cars of the 200 series manufactured by Volvo, a Swedish automobile manufacturer. The cars were imported to Puerto Rico by Trébol Motors Distributor Corporation (“Trébol Distributor”), and sold for the most part by Trébol Motors Corporation (“Trébol Motors”). Both of these companies (together “Trébol”) were owned by Jorge Gonzalez and, after his death, by his widow and son, defendants Conchita Navarro de Gonzalez and Ricardo Gonzalez Navarro (together “the Gonzalez defendants”).

Plaintiffs have advanced various arguments at different times, but the claims coalesce around one general complaint: that Volvo, Trébol, and the Gonzalez defendants conspired to violate, and did in fact violate, RICO by engaging in hundreds of predicate acts of mail and wire fraud. 18 U.S.C. §§ 1341, 1343, 1962. The plaintiffs allege that the acts were part of a complex and multifaceted scheme to defraud Puerto Rican purchasers by overcharging them for series 200 Volvo cars.

The plaintiffs’ first complaint was filed in the district court on June 16, 1992, and named as defendants Volvo, Trébol, and the Gonzalez defendants, and certain other companies no longer parties to the case. The parties spent nearly four years in bitterly waged discovery battles. On April 26, 1996, the district court granted the plaintiffs’ motion to file a Third Amended Complaint, which is the operative complaint in the suit. Trial was scheduled to begin before a jury on Monday, June 24,1996.

On the afternoon of Friday, June 21, 1996 — almost literally the eve of trial — Tré-bol and the Gonzalez defendants presented the district court with a joint notice consenting to entry of default “as to the factual averments of the Third Amended Complaint.” The district court entered a “final Partial Judgment on the issue of liability” against Trébol and the Gonzalez defendants on July 2.

Trial proceeded with Volvo as the sole defendant. The district court had earlier bifurcated the proceedings, intending that the jury would produce a general liability verdict first and then, if necessary, a verdict regarding damages. Indeed, the district court envisioned two separate damages proceedings, one to determine class-wide damages, and a second (probably involving a special master) to determine the damages of individual class members. However, the presentation of the evidence was not bifurcated, and the jury heard testimony regarding both liability and damages over twenty days of trial.

On July 25,1996, the twenty-second day of trial, the jury returned a verdict finding that Volvo was liable for violating and conspiring to violate RICO’s prohibition on racketeering activity. 18 U.S.C. § 1962(c), (d). The jury also found that Volvo had used or invested income from a pattern of racketeering activity, id. § 1962(a), but that the plaintiffs had not been injured by this violation.

On July 31, 1996, the district court held a hearing on damages. Neither plaintiffs nor Volvo presented further evidence, but both presented arguments to the jury based on the evidence already submitted. The next day, the jury found Volvo liable for $43,197,-100. The district court trebled the amount under 18 U.S.C. § 1964(c), yielding $129,591,-300. The plaintiffs then moved for prejudgment interest, attorney’s fees, and costs, alleging vexatious discovery conduct by Volvo. Volvo moved for judgment as a matter of law or a new trial.

On August 29, 1996, the court held a hearing on damages as to Trébol and the Gonza lez defendants. On October 10, 1996, the district court issued an opinion and orders denying Volvo’s motion for judgment in its favor or new trial and the plaintiffs’ motion for prejudgment interest. The court also entered a final judgment as to all defendants (including Trébol and the Gonzalezes) in the full amount of the trebled verdict. Volvo renewed its motion for judgment or new trial, and Trébol and the Gonzalez defendants filed separate motions to set aside the judgment under Fed.R.Civ.P. 59, all of which were eventually denied.

On March 27, 1997, the district court entered two lengthy orders. The first order sanctioned all defendants for vexatious conduct during discovery by declaring that the court would view plaintiffs’ submission regarding computation of attorney’s fees with indulgence. The second order awarded attorney’s fees and costs to the plaintiffs in the amount of $3,518,844.41. Trébol Motors and Trébol Distributor had filed for bankruptcy under Chapter 11 nearly six months before, on September 30,1996.

Volvo now appeals the denial of its motion for judgment as a matter of law, for a new trial, or for a remittitur (No. 97-1135), as well as the district court’s ruling on sanctions and attorney’s fees (No. 97-1599). 2 Trébol and the Gonzalez defendants appeal from the entry of judgment against them (Nos. 97-1140, 97-1143), and the Gonzalez defendants also appeal the sanction and fee award (No. 97-1790). Finally, the plaintiffs have filed a protective appeal asking that, in the event of a remand, this court should require the district court to award them prejudgment interest (No. 97-1145).

In this opinion, we address only Volvo’s appeal on the merits; the remaining appeals are dealt with in two separate opinions also issued today. In the instant appeal, Volvo argues that judgment should be entered in its favor because several elements necessary to the plaintiffs’ claims were not established by the evidence. . Volvo also makes several attacks on procedural and evidentiary rulings, which it claims warrant a remand for a new trial. 3

Volvo’s main contention is that the evidence does not support a finding that Volvo committed predicate acts of mail or wire fraud, a crucial element of plaintiffs’ RICO claim. See Ahmed v. Rosenblatt, 118 F.3d 886, 888 (1st Cir.1997). 4 We review the district court’s denial of Volvo’s motion for judgment as a matter of law de novo as to issues of law. We resolve credibility issues and draw inferences in favor of the jury verdict and examine the evidence as a whole. Still, the verdict cannot stand if no reasonáble reading of the evidence will support it. See Schultz v. Rhode Island Hosp. Trust Nat’l Bank, N.A., 94 F.3d 721, 731 (1st Cir.1996).

To prove mail or wire fraud, plaintiffs must show three elements: a “scheme to defraud,” Volvo’s “knowing and willful participation in the scheme with the intent to defraud,” and the use of the mails or interstate wire or radio communication in furtherance of the scheme. United States v. Cassiere, 4 F.3d 1006, 1011 (1st Cir.1993). The conduct must “be intended to deceive another, by means of false or fraudulent pretenses, representations, promises, or other deceptive conduct.” McEvoy Travel Bureau, Inc. v. Heritage Travel, Inc., 904 F.2d 786, 791 (1st Cir.1990), cert. denied, 498 U.S. 992, 111 S.Ct. 536, 112 L.Ed.2d 546 (1990). In addition to demonstrating the fraud, however, RICO requires the plaintiffs to show that they were “injured in [their] business or property by reason of’ the fraud. 18 U.S.C. § 1964(e).

The plaintiffs presented the jury with “five frauds” in which they claimed Volvo participated. These are not five isolated acts of fraud, but different groups of activity, each comprising multiple acts; if the activities within the group were fraudulent, any one group would reflect multiple “predicate acts” under RICO. We must determine whether the evidence presented as to at least one of the five frauds is sufficient to support the verdict. We conclude that none of the frauds has been established against Volvo, that the judgment against it must be reversed, and that plaintiffs’ claims against Volvo fail in their entirety.

Excise Taxes. First, the plaintiffs claim that Volvo and Trébol together fraudulently underreported material data to the Puerto Rican tax authorities, resulting in reduced excise tax payments on vehicles shipped to Puerto Rico. During the class period, the excise tax on a car imported to Puerto Rico was a function of its horsepower, weight, and FOB (i.e., factory-gate) price from the manufacturer. Volvo certified to Trébol the weight and horsepower of each car shipped to Puerto Rico, and Trébol then disclosed this information to the tax authorities.

Plaintiffs claim that Volvo used a computation formula for horsepower that understated the real horsepower and thus lowered the amount of excise tax due. According to plaintiffs’ expert on tax issues, Puerto Rico regulations prior to March 1989 required that horsepower be certified in accordance with standards of the Society of Automotive Engineers. It is common ground that the Society had promulgated two standards of horsepower measurement known as J245 and J1349 and that Puerto Rico’s regulations did not specify which method was to be used until March 1989, when the regulations were amended to require the use of J1349.

The undisputed evidence is that Volvo’s certifications to Trébol until March 1989 used J245, which yielded a measurement of 107 hp for 200 series vehicles. After the regulations were amended, Volvo began using J1349, which produced a measurement of 114 hp and materially increased the excise tax owed on the vehicles. As far as the record shows, Volvo’s certifications disclosed the method used in calculating horsepower and were accurate under the methods used.

Plaintiffs make a similar claim of inaccuracy as to car weight. Plaintiffs’ expert admitted that, prior to 1989, Puerto Rico had no regulations defining how a car’s “curb weight” was to be calculated. Volvo’s certifications specified that certain items were not included in the weighing, such as fluids, air conditioning, air bags, and other accessories. The weight regulations were amended in 1989 to specify that curb weight included the weight of fuels and lubricants, among other items. There is no evidence that Volvo’s certifications thereafter failed to report the higher weights as-required under the amended regulations.

Plaintiffs call Volvo’s behavior prior to 1989 “excise tax fraud.” Their contention is that Volvo knew of the alternative measurement methods for weight and horsepower, but used “false” horsepower and weight figures that enabled Trébol to “evade” payment of over $20 million dollars in excise taxes. According to plaintiffs, it is fraudulent behavior to take advantage of loopholes in the tax laws and a reportedly lax regulatory environment. 5 Plaintiffs cite no authority for the proposition that reducing one’s tax burden with methods that are not prohibited by law constitutes tax fraud. We are confident that none can be found.

Plaintiffs attempt to strengthen their argument by demonstrating that Volvo and Tré-bol knew that other methods of calculating horsepower and weight were available. In particular, they note that the owners’ manuals for similar cars sold in the United States and Canada stated horsepower according to the J1349 method well before 1989. This does not establish that Volvo or Trébol had a duty to use J1349 figures in compliance with Puerto Rico’s tax laws. Plaintiffs’ own expert stated that they did not have such a duty.

A separate issue concerns the disclosure of the FOB price. Plaintiffs proffered several submissions of Trébol Motors to the tax authorities in which the FOB price reported is substantially lower than the invoice price charged by Volvo. A jury could certainly infer that, absent an explanation, such un-derreporting — at least if systematic — constituted tax fraud by Trébol. However, no evidence implicates Volvo in false reporting of the FOB price to the Puerto Rican Bureau of Excise Taxes. The plaintiffs’ tax expert agreed that the duty to pay the excise tax applies to the Puerto Rican importer, not to Volvo. Another expert agreed that he had seen no evidence that Volvo had participated in tax evasion involving FOB prices. 6

Plaintiffs place great weight on a single sentence in a single piece of correspondence in 1987 by a Volvo representative to Trébol. The letter concerns a reduction of the length of the factory warranty for 1988 vehicles in light of Trebol’s inability to keep warranty claims within Volvo’s suggested budget. The letter points out that retaining a three-year warranty would require Volvo to raise the vehicle price by $500, whereas reducing the warranty to one year would enable Volvo to reduce the price by $500. The letter then states: “Also keep in mind that there is a hidden potential profit for you due to reduced excise taxes.”

The plaintiffs argue that this sentence refers to Volvo’s and Trebol’s joint efforts to circumvent excise taxes by underreporting data to Puerto Rican authorities. Although creative, this reading is utterly implausible because the letter, in all other respects, relates to the substantial warranty expenses incurred by Trébol. The quoted sentence can only refer to the fact that a reduction in warranty^ by reducing the invoice cost of the car, would by the same token reduce Trebol’s excise tax liability, which was calculated partly as a function of invoice cost.

Even if plaintiffs’ evidence were sufficient, they have failed to prove that they were injured “by reason of’ such activity. 18 U.S.C. § 1964(e). Since excise taxes ordinarily increase the cost to consumers, plaintiffs presumably benefitted from a reduction in taxes. They argue that, by paying lower taxes than competing automobile businesses, Trébol gained an advantage allowing it to dominate the European car market in Puerto Rico and charge higher prices than a truly free market would allow. It is uncertain whether this sequence, even if proved, would involve injury too remote to support recovery under RICO. See Holmes v. Securities Investor Protection Corp., 503 U.S. 258, 268-74, 112 S.Ct. 1311, 117 L.Ed.2d 532 (1992).

However this legal issue might be resolved, plaintiffs have pointed to no evidence in the record that other car importers bore higher tax burdens, that these taxes were a reason for a lack of competition among European cars, or that plaintiffs bought Volvos because of a lack of European-car alternatives. We therefore conclude that no rational jury could have found that the supposed “excise tax fraud” comprised acts by Volvo in violation of RICO or caused a recoverable injury to plaintiffs.

Port-Added Emblems and Accessories. Plaintiffs’ next allegation concerns emblems and accessories. During the class period, 200 series Volvo cars sold in the mainland United States and Puerto Rico had emblems affixed to them that indicated the ear model. The vehicles at issue in this case all bore emblems that identified them as 240 DL, 240 GL, or 240 GLE. The middle number “4” in the 240 model series indicates a ear with a four-cylinder engine. The underlying 240 series car ‘^platform” is the same in all eases; the evidence at trial showed that the emblems reflect varying levels of accessories. In any given market in any given year, GLEs have more accessories and are more expensive than GLs, which in turn have more accessories and a higher price than DLs. 7

The dispute centers on how the emblems and accessories came to be part of the cars as sold to Puerto Rico customers. Between 1985 and 1991, all 200 series cars sent to Puerto Rico arrived from Sweden as 240 DLs with DL emblems. Trébol, with Volvo’s permission, equipped some DLs with additional Volvo accessories and new GL or GLE emblems, all of which had been shipped separately from Sweden. The cars were then sold as either GLs or GLEs. 8 This contrasts with Volvo’s practice in most other destination markets, where vehicles were imported with accessories and emblems already factory-installed in Sweden.

Trébol also marketed a car called a “240 DLWA,” which bore a DL emblem but had certain added accessories not included in Trebol’s 240 DL. DLWA appears to be an internal designation adopted by Trébol. Although this model type was not reflected on an emblem, it did appear in Trebol’s internal price lists and some customer purchase contracts. Thus, even within a single model year, vehicles bearing the same emblem might have different accessories.

The plaintiffs claim that Volvo’s authorization of port installed accessories and emblem changes gave rise to “fake” Volvo models, permitting Trébol to sell cars that were “in reality” DLs as if they were DLWAs, GLs or GLEs commanding higher prices. Several plaintiffs testified that they felt “cheated” when they discovered that their DLWA, GL or GLE was not “from the factory” but rather was shipped from Sweden as a DL, with accessories added in Puerto Rico. Plaintiffs do not dispute that the cars and all their equipment, including the accessories installed in Puerto Rico, had been manufactured by Volvo in Sweden.

At trial, Volvo’s representative offered an arguably plausible business explanation for the practice. 9 However, Volvo apparently conceded that local installation of major accessories was a rare practice in Volvo’s other markets and did not occur in the larger continental U.S. market. And doubtless a jury might initially think that there was something suspicious about upgrading the emblem after the car left the factory.

But looked at objectively, the plaintiffs received the cars they ordered — Swedish-made Volvo 240 series car platforms with specified accessories — and this is so whether the Swedish voltmeter and upgraded tires were installed in Sweden or Puerto Rico. The Puerto Rico customers all knew what accessories their respective cars contained. The question, then, is whether Trebol’s sale of the car with no affirmative false statement as to where the accessories were installed could be treated as fraud. At the core of the issue is how fraud is to be defined.

Putting aside the scienter element in fraud and looking only at behavior, the locus classicus of fraud is a seller’s affirmative false statement or a half truth, i.e., a statement that is literally true but is made misleading by a significant omission. See Emery v. American Gen. Fin., Inc., 71 F.3d 1343, 1348 (7th Cir.1995) (Posner, C.J.). At common law, fraud doctrine did not impose any broader, general duty to disclose, see Chiarella v. United, States, 445 U.S. 222, 228, 100 S.Ct. 1108, 63 L.Ed.2d 348 (1980), but it is settled that the mail and wire fraud statutes go somewhat beyond the common law, see McEvoy Travel Bureau, Inc. v. Heritage Travel, Inc., 904 F.2d 786, 791 (1st Cir.1990). Thus, a leading commentary on federal jury instructions says that even where there is no falsehood or half truth:

[T]he failure to disclose information may also constitute a fraudulent representation [under the mail and wire fraud statutes] if the defendant was under a legal, professional or contractual duty to make such a disclosure....

2 Sand, Siffert, Loughlin & Reiss, Modern Federal Jury Instructions, ¶ 44.01, at 44-11 (1997).

There is no basis here for imputing to Volvo or Trébol a “legal, professional or contractual duty” to announce where the emblem and accessories were installed. And Judge Sand’s formulation clearly (and correctly) implies that, outside of these categories, there is no general obligation of the seller to tell the buyer everything negative that the buyer might be interested in learning about the transaction or item to be purchased. It is safe to say that few sellers do so — without an explicit regulatory requirement — and caveat emptor remains the general rule (absent false statements or half truths). 10

No doubt a shadowy area exists in which conduct alone, in context, can amount to a misrepresentation. Imagine the sale of a 240 DL where the salesman made no formal representations about the car, although knowing it to have a four-cylinder lawnmower engine in place of a car engine; or, consider a customer who explicitly told the salesman that she was buying the car because it had built-in child safety seats, when the salesman knew that the seats had been removed from that particular ear. If such deceptions were deliberate, we assume, at least arguendo, that the sales, could be treated as artifices to defraud under the federal statutes.

But nothing inherent in the process of buying a Volvo car amounts to an implicit representation that a Volvo-supplied volt meter or power antenna was necessarily installed by Volvo in Sweden. Many cars today are commonly made entirely in countries other than the home of the manufacturer, and various components are made for, rather than by, the car manufacturer. Nor does it matter that for some other markets Volvo installed more of the accessories in its Swedish factories; whether or not the jury accepted Volvo’s explanation for its practice (see note 9 above), nothing obliged Volvo to operate in the same manner in every market.

Curiously, there is not much direct precedent on the question of non-disclosure, and nothing is as helpful as the Sand formulation. 11 It would be a truly revolutionary change to make a criminal out of every salesman (assuming use of the mails or telephone) who did not take the initiative to reveal negative information about the product and who — a jury might find — harbored in his heart the hope that the buyer would never ask. Whether or not this would bring business in the United States to a standstill might be debated; but it is a change, if it is ever to be made, that should be undertaken by legislatures and not by courts. Cf. McEvoy Travel, 904 F.2d at 791-91.

The plaintiffs say that the accessories added in Puerto Rico disproportionately increased the sale price of the vehicles. They claim (for example) that the difference in cost to Trébol between a 1987 DLWA and a 1987 DL was $155, but the difference in purchase price to the customer was $3100; in other words, Trébol marked up the accessories enormously. Plaintiffs also say that the GLE label itself effectively represented a $2000 increase in price. Certain plaintiffs testified that they would not have paid these surcharges had they known that the added accessories were so minor.

However, the plaintiffs have not shown that the vehicle price or the nature of the car’s luxuries were unknown to them. No plaintiffs testified that they believed the GLE and GL were different from the DL in any way other than accessories and emblems; generally, plaintiffs signed a “Pre-Delivery Inspection Report” upon delivery of the vehicles acknowledging the accessories installed in the car. Plaintiffs’ statements in retrospect that they would rather not have paid their bargained-for price for accessories they knew they were getting are not proof of fraud. 12

Prices. The plaintiffs’ third claim is denominated as one of “price fraud.” This claim consumed much of the trial with expert testimony as to whether Volvo cars sold by dealers in the mainland United States carried substantially lower prices than equivalently equipped cars sold by Trébol in Puerto Rico. There was also documentary evidence tending to show or negate Volvo’s awareness of the retail prices of ears in Puerto Rico.

Assuming that Volvo was aware that Tre-bol’s retail prices exceeded those of comparably equipped Volvo cars on the mainland, such a price disparity is not itself fraud. An excessive price can be the result of fraud, the defrauded buyer generally paying a higher price than he would have paid absent a deception. But there is nothing in the law of fraud that prevents even a single seller from charging different markups in different markets so long as there is no affirmative misrepresentation.

However, plaintiffs’ claim does not rest solely on the premise that Trebol’s prices were higher than mainland prices. First, the plaintiffs assert that Volvo and Trébol interfered with the ability of Volvo dealers in the United States to sell cars at lower prices in Puerto Rico, and that as a result Trébol was able to keep charging excessive prices. 13 Second, plaintiffs argued that Trébol, with Volvo’s permission, discouraged buyers from purchasing vehicles in the continental United States and importing them to Puerto Rico by refusing to honor the imported vehicle’s warranty for more than one year. 14

Putting aside Volvo’s own responsibility for Trebol’s pricing, large markups and differential pricing may seem unfair to consumers and may well have appeared reprehensible to the jury. And actions taken by Volvo or Trébol or both to reinforce this market segmentation and to discourage intra-brand competition by other mainland-based Volvo dealers may have looked even worse. Indeed, some kinds of constraints on intra-brand competition are vulnerable under the antitrust laws. See Continental T. V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 58-59, 97 S.Ct. 2549, 53 L.Ed.2d 568 (1977); Packard Motor Car Co. v. Webster Motor Car Co., 243 F.2d 418, 420 (D.C.Cir.), cert. denied, 355 U.S. 822, 78 S.Ct. 29, 2 L.Ed.2d 38 (1957).

The ability to charge excessive markups and engage in differential pricing is almost everywhere the symptom of lack of competition or other market failure (e.g., inadequate information). Such behavior flourishes in markets that are small and are geographically or otherwise isolated. For products like automobiles, customers tend to shop where they can also get service and one expects less competition and higher markups in thinner markets (where costs may also be higher). Whether price discrimination helps or hurts consumers, and it can do either, VIII Areeda, Antitrust Law ¶ 1616f, at 226 (1989), it is not itself fraud.

We have no reason to consider whether the antitrust laws would reach Volvo’s discouragement of Florida dealers from marketing their cars in Puerto Rico or Volvo’s use of warranty policies to dissuade customers from buying their cars elsewhere and bringing them to Puerto Rico. Antitrust violations were not charged in this ease, and in any event do not constitute predicate acts under RICO. 18 U.S.C. § 1961(1). This is also not the rare case where an antitrust violation also involves a sufficient measure of deception to qualify as fraud, {e.g., secret bid-rigging cases).

Double Invoicing. The most factually complex claim against Volvo arises from a double invoicing scheme involving a Liechtenstein-based corporation named Auto und Motoren Aktiengesellsehaft (“AUM”). Tré-bol, Volvo, and AUM maintained what was, according to Volvo, a relationship whereby AUM guaranteed Trebol’s debts to Volvo. Volvo would send Trébol an invoice with every shipment of vehicles and copy that invoice to AUM to keep it informed of its potential liability.

Plaintiffs showed that, between 1986 and 1989, AUM sent Trébol additional invoices for the same vehicles but with different purported prices. AUM’s invoices sometimes overstated the vehicles’ cost relative to the actual price in the authentic Volvo invoice by as much as $3000; not all AUM invoices contained inflated prices and some of the inflated ones involved small amounts {e.g., $50). In any case, Trébol used these inflated invoices to obtain higher inventory financing from Puerto Rican banks and Tré-bol also calculated its retail prices using the inflated cost as a starting point. Trébol also remitted the higher invoice amount to AUM, resulting in the accumulation of a pool of excess funds in Liechtenstein. Plaintiffs’ evidence suggested that some of these funds were funnelled back to Trébol.

Since Trébol was sent copies of the original Volvo-prepared invoices, a rational jury could conclude that Trébol knew of the actual prices and was defrauding Puerto Rican banks by means of the double invoicing scheme. If Trébol used the inflated invoice prices on its tax returns, tax authorities may also have been defrauded. Far more tenuous is the conclusion that Volvo knowingly participated in the fraud. After reviewing the evidence presented, we do not believe that any rational jury could conclude that it was more likely than not that Volvo knowingly participated in the AUM fraud.

Plaintiffs say that an inference of such knowledge could be inferred because Volvo’s relationship with AUM was unusual and not adequately explained. Plaintiffs point out that Volvo importers in other markets usually did not invoice through a foreign guarantor. Plaintiffs’ expert testified that AUM served no legitimate business purpose: he said that it seemed to serve simply as a conduit for Trebol’s payments and despite the “guarantor” label, in early 1993, Volvo executed a “Loan Agreement” whereby Volvo loaned AUM 2.7 million dollars to cover Trebol’s debts — a transaction that appears exactly contrary to a guarantor’s role. Volvo later authorized a transaction in which AUM swapped Trebol’s debt to it for equity in Trébol.

The use of a “guarantor” for Puerto Rico but not for the continental United States is less mysterious than might otherwise be supposed. The evidence showed that, in the continental United States, Volvo’s importer was a wholly owned subsidiary of Volvo and the use of a guarantor would have been superfluous. In Puerto Rico, Volvo was entrusting valuable inventory to an importer controlled by the dealer and in a location where Volvo had no permanent presence.

Why Volvo would have lent money to a guarantor on one occasion was not as well explained; Volvo’s representative described the transaction as involving no real loan of money but as an “extended credit term” to give Trébol an extension on its payments for cars shipped by Volvo. One might have expected Volvo, in conducting a prudent defense, to have shed more light on its connections with AUM, assuming that they were wholly innocent. But the burden was not on Volvo to negate charges of fraud, and the trial judge mistakenly hindered efforts by Volvo to provide some explanation or negate the inference. 15

In all events, on the evidence of record, it would be pure speculation for a jury to conclude that Volvo had any knowledge of the double invoice scheme. Self-interest gave Trébol reason to participate in such a fraud, but there was little reason for Volvo to do so absent proof — and proof was absent — that Volvo got more than its original invoice price. The jury had no basis for concluding that Volvo received more money from AUM than was charged in the original invoices. 16

Plaintiffs rely on United States v. London, 66 F.3d 1227, 1243 (1st Cir.1995), in which we held that “evidence of mutual interest” between a money launderer and his clients supported an inference that the launderer “intended to promote” his client’s illegal activities. However, the inference in London was permissible only because “evidence of mutual interest” was coupled with “know[ledge] that the property involved ... represented] the proceeds of some form of unlawful activity.” Id. at 1242 (quoting 18 U.S.C. § 1956(a)(1)(A)(i)); see also United States v. Ortiz,. 966 F.2d 707, 713-14 (1st Cir.1992), cert. denied, 506 U.S. 1063, 113 S.Ct. 1005, 122 L.Ed.2d 154 (1993). Plaintiffs have not shown analogous knowledge on Volvo’s side that Trébol was engaging in fraudulent activity.

Disclosure Labels. The last of the “five frauds” involves misrepresentations of information on disclosure labels required by law. During most of the class period, a Puerto Rican statute, popularly known as “Law 77,” mandated that the “seller or dealer” of motor vehicles conspicuously affix to each vehicle for sale a label disclosing, inter alia, the name and address of the selling entity, the factory cost of the vehicle, the tax “on the vehicle and equipment installed in Puerto Rico,” and the “[r]etail price in Puer-to Rico suggested by the seller.” P.R. Laws Ann. tit. 23, § 1023 (1987), repealed by 1992 P.R. Laws Act 80. 17

Plaintiffs introduced several automobile disclosure labels showing “cost” figures that were based on the inflated figures sent by AUM. Because the information actually provided in Trebol’s labels is critical to this claim of fraud, we have reproduced an example of such a label in the appendix to this opinion, omitting specific cost numbers. Although Trebol’s labels did not precisely follow the requirements of Law 77, it is clear that they were intended to be Law 77 labels.

Trébol appears to have begun with AUM’s inflated CIF (“cost, insurance, freight”) figure and added more items — occasionally itemized by Trébol in the margin of AUM invoices as “royalties,” “accessories,” “publicity,” and “other costs” — to produce a base “cost” figure .that was reported on the Law 77 label. Trebol’s Law 77 label then added excise taxes, license plate fees, compulsory insurance and “other costs” to produce a “total vehicle cost.” The label also stated a “suggested sales price” which was somewhat higher {e.g., $1,500) than the “total vehicle cost”; actual purchase prices generally fell between the total vehicle cost figure and the suggested sales price.

To the extent that the AUM invoices gave inflated CIF cost figures (and not all did), the Law 77 labels that copied those'figures misrepresented what Law 77 intended to be the factory “cost” of the vehicles to Trébol. The jury could easily infer that this misrepresentation injured some of the plaintiffs; in bargaining with Trébol, customers probably believed that the dealer had to charge at least the reported “total vehicle cost.” Deliberately overstating this figure would be fraud, and we need not look far to find a use of the mails in connection with the scheme, namely, Trebol’s correspondence with-AUM which furnished the inflated CIF figures.

However, plaintiffs do not dispute that the label provision of Law 77 imposes the disclosure obligation only upon sellers or dealers, which are clearly distinguished from “manufacturers” in the statute. See P.R. Laws Ann. tit. 23, §§ 1022, 1023. There is no evidence, nor is there a basis to infer, that Volvo had any role in the preparation or display of the labels. Indeed, as already explained, there is no evidence that Volvo ever knew that AUM was furnishing false figures to Trébol, which Trébol then used in the Law 77 labels.

Alternatively, plaintiffs argue that Volvo must at least have been aware that false labels were being affixed. Its representatives, they say, visited Trébol for quarterly meetings to discuss marketing and sales. At least one Volvo witness said he was familiar with Trebol’s prices to customers, although he did not say that he learned this information from looking at the labels. Volvo was also entitled under its importer agreement to direct reports from Trébol as to prices and sales.

We will assume arguendo (and seemingly without any direct evidence) that Volvo representatives saw the Law 77 labels, that they read Spanish, 18 and that they took note not only of the suggested sales price on the labels but the underlying cost figures as well. As álready noted, Law 77 requires a statement of the “factory cost of the vehicle.” If the labels had been phrased in this way, perhaps they would have been alerted to a discrepancy between the ordinary factory price of an unequipped DL and the higher figure on the label derived from the false AUM invoices.

But the actual labels did not refer to “factory cost” at all, although Law 77 intended that this figure be furnished; instead, the base figure stated on the label is identified as “costo (incluye flete y seguro),” ie., cost including shipping and insurance. There is no indication on the label whether the cost purports to be that of the manufacturer or of the importer — remember, here the “seller” is third in the chain — or whether it includes the accessories shipped separately from Sweden, or whether a potentially significant importer markup is included anywhere in the cost figure.

To infer from this that Volvo knew that Trébol was using false Law 77 labels is a leap that cannot rationally be made. Plaintiffs do not offer any persuasive explanation as to precisely how the Law 77 labels must have revealed fraud to Volvo. Nor do provisions of Volvo’s standard importer agreement granting Volvo the right to request reports on Trebol’s sales and; financial situation suffice; not only is it unclear what information Volvo received, but there is no showing that Trebol’s sales prices and balance sheets would have revealed the “cost” misrepresentations on Law 77 labels. 19

Turning from Trebol’s disclosure obligations to those of Volvo, plaintiffs argued Volvo itself was obliged to ensure that the cars bore a different label, required under the Automobile Information Disclosure Act (commonly called the “Monroney Act”), 15 U.S.C. §§ 1231-33. This statute requires that every “manufacturer of new automobiles distributed in commerce” in the United States, including Puerto Rico, have a label affixed to the windshield or side window pri- or to delivery to the dealer, specifying information akin to, but different in some respects from, the disclosure required under Law 77.

The statute defines “manufacturer” as “any person engaged in the manufacturing or assembling of new automobiles, including any person importing new automobiles for resale.” 15 U.S.C. § 1231(a). Volvo asks us to read the Monroney Act as if the duty to install the label on imported cars falls on the importer rather than upon the foreign manufacturer. This suggestion has some force— merely as a matter of language — because the label requirement is directed to the affixing of the label “prior to the delivery of any new automobile to any dealer,” id. § 1232, which is the task usually .performed by the importer in the case of a foreign car.

Volvo’s interpretation is also supported by some legislative history which speaks of disclosure by the manufacturer “or” importer, see H.R.Rep. No. 85-1958 (1958), reprinted in 1958 U.S.C.C.A.N. 2902, 2903, 2906, and by the fact that certain of the information required in the label would be known to the importer but not necessarily to the manufacturer, see, e.g., 15 U.S.C. § 1232(c) (name and address of each dealer). Two trade associations have filed an amicus brief asserting that the uniform practice in the case of foreign car imports is for the importer rather than the manufacturer to affix the label.

We were unable to locate any decided case on this issue and need not resolve it here. As plaintiffs themselves assert, Volvo was not sued under the Monroney Act, and the Mon-roney Act is not one of the statutes whose violation comprises a predicate act under RICO. 18 U.S.C. § 1961(1). Even if the Monroney Act did apply to Volvo, plaintiffs’ fraud claim on this theory would fail for lack of proof of scienter: as already noted, the offense of fraud requires both a deceptive act and knowledge that the act is deceptive. Cassiere, 4 F.3d at 1011.

A mail or wire fraud conviction usually cannot result from a failure to disclose unless the defendant was aware of its duty to disclose. See, e.g., Cassiere, 4 F.3d at 1022; Sand, supra, ¶ 44.01, at 44-11, 44-16. If Volvo knew it was required to disclose information under the Monroney Act and failed to do so, such a deliberate violation could constitute an act of fraud. See, e.g., United States v. Melvin, 544 F.2d 767, 774 (5th Cir.1977). However, if the disclosure was omitted in good faith, there was no fraud. United States v. Dockray, 943 F.2d 152, 155 (1st Cir.1991) (holding that good faith is an “absolute defense” to mail and wire fraud); United States v. Ashman, 979 F.2d 469, 480 (7th Cir.1992) (same).

Volvo’s duties under the Monroney Act are far from clear, and plaintiffs offered no evidence that Volvo knew it had a duty to affix labels to cars sold by Trébol. Volvo’s practice on the mainland was to let its own importing company attach the labels, there was no showing that foreign manufacturers ordinarily affixed such labels, and there is no indication that any foreign manufacturer has ever been held hable for failing to affix a label. Under these circumstances, an inference of deliberate deception by Volvo would be impossible to sustain even if we agreed that Volvo had such a legal duty under the Monroney Act — a point on which we are at the very least doubtful.

To conclude, it is only with great reluctance that an appeals court rejects a jury verdict for lack of evidence to support it. But save for acquittals in criminal cases, the right to jury trial under our Constitution does not mean that the jury automatically gets the last word: a jury decision is final only where a jury, operating within the framework of the governing legal rules, could rationally decide the case in the manner it did. And the error may not be the jury’s fault at all.

In this ease, we do not know whether the jury was misled by jury instructions whose validity we have not had to consider, or whether it wrongly attributed Trebol’s fraud to Volvo or found offensive the higher markups charged in Puerto Rico, or whether the explanation lies somewhere else. It is enough here that the evidence was not sufficient, under a proper reading of the law, to impose liability on Volvo. Volvo is entitled to judgment as a matter of law on the merits of the complaint.

The district court’s judgment entered on October 10, 1996, is reversed as to Volvo.

It is so ordered.

APPENDIX

TREBOL MOTORS

DECLARACION DE INFORMACION AL CONSUMIDOR DE ACUERDO A LOS REQUISITOS DE LA LEY NUMERO 77 APROPADA EL 31 DE MAYO DE 1972

1. NOMBRE DEL VENDEDOR

DIRECCION DEL LOCAL

DIRECCION POSTAL

2. NOMBRE DEL IMPORTADOR

DIRECCION DEL LOCAL

TREBOL MOTORS CORP.

PO BOX 11204 FDZ JUNCOS SANTURCE, PR 00910

' 3. FECHA INTRODUCCION 10/11/87

MARCA VOLVO

NUMERO DECLARACION 88041

MODELO 4DLWA

AND 1998

NUMERO DE MOTOR YV1AX8845J1285905

4 PRECIO DE VENTA SUGERIDO ******************************

5 PRECIO CONTRIBUTIVO **************************************

6. COSTO (INCLUYE FLETE Y SEGURO) ************************

IMPUESTOS (ARBITRIOS) *************************************

DERECHOS DE REGISTRO (TABILLAS) ***********************

SEGURO COMPULSORIO (ACAA) ******************************

OTROS COSTOS ***********************************************

TOTAL COSTO VEHICULO ^.**íi;^^^H;****************************

STOCK: 4DLWA3090

1 . The next opinion is directed to the principal appeals of the remaining codefendants (Nos. 97-1140 and 97-1143) and a conditional cross-appeal by the plaintiffs (No. 97-1145). The final opinion deals with sanctions and attorney’s fee 'issues (Nos. 97-1599, 97-1600 and 97-1790).

2 . The latter rulings are also the subject of a separate appeal by Volvo's attorneys (No. 97-1600) out of a concern that they may be implicated in the sanctions and award of attorney’s fees.

3 . Specifically, Volvo claims the district court abused its discretion in certifying the plaintiff class, in sanctioning Volvo’s attorneys in the presence of the jury and without notice or a hearing, in admitting certain expert testimony, and in admitting evidence of Volvo’s reputation. Volvo also criticizes the district court’s charge to the jury. For reasons that will become apparent, we do not reach any of these issues.

4 .18 U.S.C. § 1962(c) prohibits the conduct of an enterprise's affairs through a “pattern of racketeering activity” and this in turn requires "at least two acts of racketeering activity” within 10 years of each other. 18 U.S.C. § 1961(5). Racketeering activities include mail and wire fraud. Id. § 1961(1)(B).

5 . The plaintiffs' expert, a former director of the Bureau of Excise Taxes at Puerto Rico’s Department of the Treasury, stated on cross-examination that tax forms were sometimes processed hy government employees unfamiliar with the regulations and that the government relied on automobile importers to report the measurements according to the less favorable standard.

6 . Plaintiffs suggest that Volvo nonetheless facilitated Trebol’s understatement of FOB prices by including a price increase as a "separate line item" on invoices at Trebol's request. Plaintiffs argue that the separate surcharge was not included by Trébol in the taxable FOB price reported to the tax authorities. Yet the surcharge was clearly indicated on Volvo’s invoices and would have been evident to the tax authorities if disclosed. Trébol did not submit the invoices to the authorities, and it is unclear whether it had a duty to do so. But it is clear that the price Trébol submitted for tax purposes was not controlled by Volvo’s itemization of price elements on the invoices.

7 . Volvo offered undisputed testimony that the emblem designations were not comparable across different markets or years, so that a DL in the United States might have more accessories than a .GLE in Sweden in the same year, and ■ a DL in Puerto Rico in one year could have more accessories than the GL of the previous year.

8 . Some of the accessories installed at the port in San Juan included more ornate tires, metallic paint, leather interiors, voltmeters, tachometers, aluminum wheels, power antennas, and additional radio speakers.

9 .A Volvo officer testified that Trébol had asked Volvo to provide cars with more accessories than the 240 DLs and GLs that Volvo had been sending to the mainland United States and Puerto Rico. Volvo's testimony suggested that the car-maker found it inefficient for its production facilities in Sweden to produce a small number of highly-accessorized cars for Puerto Rico and that it was more cost-effective to ship a basic DL to Puerto Rico and allow Trébol to customize more luxurious vehicles as the need arose.

10 . For this reason, an array of statutes and regulations have been enacted to require specified disclosures for certain kinds of transactions. The statute most closely in point, discussed below, is the Monroney Act which requires specified disclosures by an automobile dealer.

11 . There are a few cases whose general language makes clear that the federal statutes go somewhat beyond common-law fraud, but all of them recognize that there must be some artifice to defraud, in addition to an illicit intent, and, most significantly, almost all of the cases can be explained either by the existence of an independent duty to disclose, e.g., United States v. Melvin, 544 F.2d 767, 774 (5th Cir.1977), or by the use of deceptive half truths, Emery, 71 F.3d at 1347; United States v. Townley, 665 F.2d 579, 585 (5th Cir.1982).

12 . We note the similarity of the entire claim discussed in this section to an earlier claim brought by a single Volvo purchaser in Puerto Rico. This claim was dismissed by the district court, and the dismissal was in turn affirmed in an unpublished decision of this court, Rodriguez-O’Ferral v. Trebol Motors Corp., 1993 WL 261993, 998 F.2d 1001 (1st Cir.1993) (table). Although the parties make reference to our earlier affir-mance, we do not rely on that decision as precedent because it was unpublished.

13 . Specifically, a Volvo dealer from Miami offered to fly Puerto Rican purchasers to Miami for a weekend and pay for accommodations if they bought a Volvo. Trébol requested that Volvo "take action” on the matter. Although Volvo did maintain a policy of discouraging its importers from interfering in each other's markets, a Volvo representative stated that there was little that Volvo or an importer could do to prevent an independent dealer from selling outside its territory. Plaintiffs have not shown that Volvo's actions on the "gray market” issue went beyond informing the Miami dealer of its disapproval.

14 .Plaintiff Canino purchased a DL in New York and paid shipping and excise to Puerto Rico, and the' resulting price was still $4000 lower than what she would have paid for an identical DL from Trébol Motors. However, Trébol would only honor the vehicle's warranty for one year rather than three because the vehicle was purchased outside Puerto Rico. When Canino asked Volvo to explain this, Volvo replied that it could not interfere with Trebol’s warranty policies.

15 .During Volvo's cross-examination of one of the plaintiffs’ experts, the district judge sanctioned Volvo’s counsel for misconduct during discovery, particularly regarding Volvo’s delay in filing copies of the invoices that Volvo had issued to Trébol and AUM. The sanction was that Volvo "cannot bring any evidence pertaining to the double invoicing sent by Volvo to the Liechtenstein corporation.” The district court eventually withdrew its sanction order, but it denied Volvo’s request to recall witnesses whose cross-examination had been limited pursuant to the order.

16 . After a heated dispute, the district court ruled that Volvo had produced for plaintiffs its records of AUM payments to Volvo. It is not clear that plaintiffs ever offered the records in evidence and, in any event, plaintiffs point to no evidence that Volvo did receive excessive payments.

17 . In 1992, Puerto Rico enacted a new statute requiring the "importer or distributor” to affix a label to all vehicles displaying "the information the Secretary [of the Treasury] deems necessary” and the “suggested consumer sales price.” 1992 P.R. Laws Act 80, § 2.

18 . Most Volvo representatives who visited Puerto Rico were Swedish, and the evidence shows that they generally conducted business with Trébol in English.

19 . We need not reach the interesting question whether — if Volvo did know of the fraud — the facts would allow a jury to hold Volvo liable under RICO for Trebol’s deceptions simply because Volvo continued to supply cars to Trebol. Cf. Direct Sales Co. v. United States, 319 U.S. 703, 709-13, 63 S.Ct. 1265, 87 L.Ed. 1674 (1943); Aetna Cas. & Sur. Co. v. P & B Autobody, 43 F.3d 1546, 1562 (1st Cir.1994).