Beasley v. Wells Fargo Bank (1991)Annotate this Case
ALICE M. BEASLEY et al., Plaintiffs and Respondents, v. WELLS FARGO BANK, N.A., Defendant and Appellant.
(Superior Court of the City and County of San Francisco, No. 861555, Robert L. Dossee, Judge.)
(Opinion by Reardon, J., with Perley, Acting P. J., and Stein, J.,fn. * concurring.)
Heller, Ehrman, White & McAuliffe, M. Laurence Popofsky, Weyman I. Lundquist, Robert S. Venning, Brian P. Brosnahan and Sarah Rubenstein for Defendant and Appellant.
Sturdevant & Sturdevant, James C. Sturdevant and Patricia Sturdevant for Plaintiffs and Respondents.
In this case we affirm a $5,227,617 judgment in a class action which challenged Wells Fargo Bank's assessment of fees against credit card customers who failed to make timely payments or exceeded their credit limits. [235 Cal. App. 3d 1389]
We hold as follows: (1) although there is no right to a jury trial on an affirmative claim for relief from liquidated damages, a jury trial was proper here because the plaintiffs sought defensive relief by opposing a cross- complaint; (2) the judge erred in submitting to the jury the legal question whether the challenged fees were valid as liquidated damages, but the bank has shown no prejudice from the error; (3) the plaintiffs could obtain monetary relief under the statute governing the validity of liquidated damages; (4) the applicable limitations period was four years; and (5) the judgment was supported by substantial evidence.
This litigation concerns two types of fees-"late" and "overlimit"-that Wells Fargo imposed on customers who had Mastercard or VISA credit card accounts. The late fee was assessed against customers who did not make a monthly payment on time. The overlimit fee was charged to customers whose account balances in a given month exceeded their credit limits by more than 15 percent.
Until 1982, the late fee was calculated as the greater of either $1 or 5 percent of the minimum payment due, but not exceed $5. The overlimit fee was $5. In March 1982, a Wells Fargo "Credit Card Task Force" proposed increasing these fees-which "working papers" described as a "good source of revenue"-pursuant to a strategy of "maximizing fee income." The person ultimately responsible for the decision to increase fees, Executive Vice-president Jack Kopec, requested a cost study by a Wells Fargo employee, Stephen Simpson. Kopec had already decided to increase the fees, but did not decide the amount of the increases until the cost study was prepared.
Simpson's cost study, completed in June 1982, recommended that the bank increase the late charge to the greater of either $2 or 10 percent of the minimum payment due, but not to exceed $10, and impose the existing $5 overlimit fee when account balances exceeded credit limits by more than 10 percent. Kopec, however, declined to follow Simpson's recommendations. Instead, he decided to increase the minimum late charge to $3 and to increase the overlimit charge to $10. The new fees became effective on December 1, 1982. The new fees, like the superseded fees, were set forth in "Customer Agreement and Disclosure Statement" forms.
On July 23, 1986, Alice M. Beasley filed a class action against Wells Fargo, seeking recovery of fees already assessed and an injunction against future imposition of fees. An amended complaint added two other named plaintiffs. The class was certified as consisting of all California customers [235 Cal. App. 3d 1390] whose accounts were assessed late and overlimit fees between July 23, 1982, and May 18, 1987. The complaint included allegations that the plaintiffs were entitled to monetary recovery under Civil Code section 1671, which governs the validity of liquidated damages, and were entitled to injunctive relief under Business and Professions Code section 17200 et seq., which proscribes unfair business practices.
Wells Fargo filed a cross-complaint for breach of contract, seeking to recover "all sums due and owing" to the bank by "certain members of the purported class" who had been assessed "certain service charges." The bank later stipulated to denial of class treatment of the cross-complaint, but the claims against the named plaintiffs remained.
Under the applicable law, Wells Fargo's late and overlimit fees were not valid as liquidated damages unless it would have been "impracticable or extremely difficult to fix the actual damage" from late and overlimit activity (Civ. Code, § 1671, subd. (d)) and the bank had made a "reasonable endeavor" to estimate a "fair average compensation" for its loss from such activity. (Garrett v. Coast & Southern Fed. Sav. & Loan Assn. (1973) 9 Cal. 3d 731, 738-739 [108 Cal. Rptr. 845, 511 P.2d 1197].) Absent either of these elements, the fee provisions in the customer agreement and disclosure statement were void, but the plaintiffs still remained liable for "the actual damages resulting from" their late and overlimit activity, which would include "administrative costs reasonably related to collecting and accounting for" late and overlimit balances. (Id., at p. 741.)
Wells Fargo sought at trial to justify its fees as liquidated damages and to prove that its actual damages exceeded the fees assessed. The bank presented expert testimony by a certified public accountant, Paul Regan, who claimed that for a representative period during the first six months of 1986, the bank's late fees were $3,212,000 while its late account costs were $4,043,082, and its overlimit fees were $374,060 while its overlimit costs were $387,453. The plaintiffs presented contrary testimony by their own certified public accountant, John Lehman, who testified that for this six-month period Regan had overstated late account costs by $1,873,896 and had overstated overlimit costs by $198,214. In accordance with Lehman's calculations, the plaintiffs argued that for the entire class period the assessed late fees exceeded late account costs by at least $8,894,988, and the assessed overlimit fees exceeded overlimit costs by at least $2,358,148.
The trial judge submitted the liquidated damages issues to a jury. In a special verdict, the jury made findings on two subissues: whether the purported liquidated damages provisions were valid, and, if not, what was the extent of the plaintiffs' liability for actual damages caused by late and [235 Cal. App. 3d 1391] overlimit activity. On the validity issue, the jury found that although it had been impracticable or extremely difficult to fix actual damages, Wells Fargo had not made a reasonable endeavor to estimate a fair average compensation for loss; thus the purported liquidated damages provisions were void. On the actual damages issue, the jury found that late fees had exceeded late account costs by $4,182,796, and overlimit fees had exceeded overlimit costs by $1,044,821. Taking into account certain adjustments the jury made to the plaintiffs' calculations of fee income, the award was approximately half the amount the plaintiffs had requested.
The court independently decided the unfair business practices claim, ruling for Wells Fargo because "the equities do not favor granting injunctive relief, nor, as a matter of policy, is this Court well suited to regulating retail bank pricing via injunction on an ongoing basis." The court rendered judgment on the jury verdict and dismissed Wells Fargo's cross-complaint without prejudice. Wells Fargo filed a timely notice of appeal.fn. 1
A. Jury Trial Issues
Wells Fargo's first claim of error concerns the propriety of submitting the liquidated damages issues to the jury. The bank contends there was no right to a jury trial in this action for relief from liquidated damages, and even if there was, the subissue of validity was still a matter to be decided by the trial judge.
1. The right to a jury trial.
 In determining whether there is a right to a jury trial in an action for relief from liquidated damages, the pivotal question is whether there was a right to a jury trial of such an action at common law as of 1850, when the California Constitution was adopted. (E.g., C & K Engineering Contractors v. Amber Steel Co. (1978) 23 Cal. 3d 1, 8 [151 Cal. Rptr. 323, 587 P.2d 1136].) This question is "purely historical" (ibid.), and in this case requires an examination of the common law of England.
Before 1697, the English courts of equity had exclusive jurisdiction to relieve a party from a penalty for breach of contract. (2 Story, Commentaries [235 Cal. App. 3d 1392] on Equity Jurisprudence (1836) § 1301, p. 536.) In 1697, Parliament enacted the "Statute of William" (8 & 9 William III, ch. 11, § 8), which provided that in an action at law to recover a penalty for breach of contract, the plaintiff could only recover actual damages from the breach of contract, which were to be assessed by the jury (XII Holdsworth, A History of English Law (1938) pp. 519-520). The Statute of William did not purport to invest jurisdiction in the law courts to grant affirmative relief from the penalty, but merely permitted relief to be sought defensively in an action at law to recover the penalty.
 Thus, at common law, a party in breach was not entitled to a jury trial in an action for affirmative relief from a penalty for the breach, over which exclusive jurisdiction remained in the equity courts. The right to a jury trial was limited to defensive relief in the form of assessment of actual damages where the nonbreaching party sued at law to recover the penalty.
Ordinarily, therefore, an action against Wells Fargo for affirmative relief from late and overlimit fees would be considered equitable, with no right to a jury trial. There being no such right at common law as of 1850, there is no such right now.fn. 2
But this litigation did not consist solely of an action by customers for affirmative relief. Wells Fargo filed a cross-complaint for breach of contract, seeking to recover "all sums due and owing" to the bank by "certain members of the purported class" who had been assessed "certain service charges." The plaintiffs filed an answer which asserted the defense that the fees were "unlawful liquidated damages." Thus, by its cross-complaint, Wells Fargo was pursuing an action at law to recover fees, to which the plaintiffs could seek defensive relief with a right to a jury trial. In effect, the cross-complaint took the litigation beyond the exclusive realm of equity jurisdiction and within the scope of legal jurisdiction, where defensive relief could be obtained from a jury. The bank later stipulated to denial of class treatment of the cross-complaint, but the claims against the named plaintiffs remained.
Wells Fargo points out that it effectively abandoned the cross-complaint entirely, even against the named plaintiffs, by submitting no jury instructions or special verdict questions on the cross-complaint. Such abandonment, however, did not become fully evident until the close of trial, long after the right to a jury trial had been litigated. The judge could not have known at the time he denied the objection to a jury trial that the cross-complaint would [235 Cal. App. 3d 1393] later be "effectively" abandoned, and thus cannot be faulted on that basis for allowing a jury trial.
In short, even though the plaintiffs did not have a right to a jury trial on their suit for relief from liquidated damages, the bank's cross-complaint and the plaintiffs' answer created a defensive claim for relief on which there was a right to a jury trial, and thus the bank's objection to a jury trial was substantively meritless.fn. 3
2. Submission of the validity subissue to the jury.
The plaintiffs' right to a jury trial was not, however, absolute. Even in actions at law before 1850, juries did not decide the validity of a liquidated damages provision-i.e., whether the sum to be paid in the event of a breach constituted liquidated damages or an unlawful penalty. This was a matter to be decided by the judge.
As stated in Sainter v. Ferguson (1849) 137 Eng.Rep. 283, 288, although the question of validity "has sometimes been left to the jury ... it is now clearly settled, that, whether the sum mentioned in an agreement to be paid for a breach, is to be treated as a penalty or as liquidated and ascertained damages, is a question of law, to be decided by the judge upon a consideration of the whole instrument." (See also McCormick, Handbook on the Law of Damages (1935) § 157, p. 624 [courts hold "almost unanimously, that the trial judge, not the jury, decides in the light of all the facts whether the stipulation is for liquidated damages or for a penalty"]; 11 Cal. Law Revision Com. Rep. (1973) 1229, 1291, fn. 291 [case law is not clear but "it appears that the question of validity will be determined by the judge"].)fn. 4
 Thus, while the amount of actual damages from late and overlimit activity was a question of fact for the jury to decide, the preliminary question whether the liquidated damages provisions were valid was a matter for the court to decide. The court, not the jury, should have decided whether it had been impracticable or extremely difficult to fix actual damages and whether Wells Fargo had made a reasonable endeavor to estimate a fair average compensation for its loss. [235 Cal. App. 3d 1394]
This conclusion raises two questions which the parties have not addressed. The first is whether Wells Fargo waived the error by failing to object to submission of the validity issue to the jury. The answer is no. Wells Fargo first argued this point almost a year before trial in a motion to strike the jury demand, which was styled alternatively as a motion "to set [the] order of proof of issues to be tried to the court." Shortly before trial, the bank again made this argument in a trial brief. Even after trial, in a motion for judgment notwithstanding the verdict, the bank argued that the court should independently decide the validity issue.
 The second question is as follows: If validity was a matter for the judge to decide, can we decide that matter independently on appeal? Again, the answer is no, but the question is more problematic.
Ordinarily, a trial judge's decision on a question of law is not binding on appeal, and the appellate court will make its own independent determination. (E.g., Parsons v. Bristol Development Co. (1965) 62 Cal. 2d 861, 865-866 [44 Cal. Rptr. 767, 402 P.2d 839] [independent appellate interpretation of a writing absent conflicting extrinsic evidence].) The problem here is that the validity issue is not really a classic question of law, but is one of fact that, because of its character, is nevertheless committed to judicial determination. One authority explains that each of the various questions pertinent to validity "is a question of fact in the sense of not being a question as to the content of a legal rule. Each of them, however, involves the application of a vague standard to a hypothetical situation. A jury is likely to make little effort to consider the imaginary picture of what could have been foreseen to the exclusion of what later actually happened, or to apply discriminatingly the standard of proportion [of agreed amount to reasonably estimable probable damages], and the courts seem wise, therefore, in holding, as they do almost unanimously, that the trial judge, not the jury, decides in the light of all the facts whether the stipulation is for liquidated damages or for a penalty." (McCormick, Handbook on the Law of Damages, supra, § 157, pp. 623-624.)
Validity being a factual issue, it would be particularly inappropriate for us to decide that issue independently if it turns on witness credibility. Appellate courts should defer to trial judge decisions "whenever the trial judge's 'nether position' in the judicial pyramid makes him a presumptively more capable decisionmaker ([Rosenberg, Judicial Discretion of the Trial Court, Viewed From Above (1971) 22 Syracuse L.Rev. 635,] 663) because of 'his observation of the witnesses [and] his superior opportunity to get "the feel of the case." ' (Noonan v. Cunard Steamship Co. (2d Cir. 1967) 375 F.2d 69, 71 (by Friendly, J.).)" (Hurtado v. Statewide Home Loan Co. (1985) 167 Cal. App. 3d 1019, 1024-1025 [213 Cal. Rptr. 712], disapproved on another [235 Cal. App. 3d 1395] point in Shamblin v. Brattain (1988) 44 Cal. 3d 474, 479, fn. 4 [243 Cal. Rptr. 902, 749 P.2d 339].)
The validity issue does indeed turn on witness credibility. The plaintiffs' theory on the "reasonable endeavor" question is that Executive Vice-president Jack Kopec made the decision to raise the fees in March 1982 in order to maximize income and generate additional revenues, and he obtained the Simpson cost study merely to justify the anticipated increases rather than as part of a reasonable endeavor to estimate a fair compensation for loss. The plaintiffs rely on the March 1982 "working papers" of the bank's credit card task force-of which Kopec was a member-which described the fees as a "good source of revenue" and proposed their increase pursuant to a strategy of "maximizing fee income."
But the evidence was in conflict on this point. Kopec testified that in stating the strategy of "maximizing fee income" and characterizing the fees as a "good source of revenue," he meant he was "trying to recover our costs," which he had assumed were no longer being recovered due to high inflation since "fees had not increased in many years and yet expenses had ...." Kopec testified that in March 1982 he did not know how much he would increase the fees; the final decision as to amount was not made until June 1982, after the preparation of Simpson's cost study. Kopec claimed he did not decide to charge merely "what the market would bear"; if he had, he would have accepted Simpson's recommendations. Kopec did not do his own cost study after Simpson's cost study was prepared, but he did do a "mental calculation."
Thus, while the March 1982 characterizations of the bank's strategy as "maximizing fee income" and the fees as a "good source of revenue," coupled with the fact Kopec had decided to raise fees before obtaining the Simpson cost study and did not do his own cost study, strongly suggest the fee increase was an attempt to increase profits rather than estimate a fair compensation for loss, Kopec's testimony was unequivocally contrary, asserting a plausible (whether or not convincing) explanation consistent with a reasonable endeavor to obtain compensation for increased costs. The reasonable endeavor issue, therefore, turned on the credibility of Kopec's testimony. As an appellate court limited to a paper record, we are ill- equipped to decide this credibility issue independently (although substantial evidence plainly supported the judgment on this point).
In short, the trial court erred in submitting to the jury the subissue whether the late and overlimit fees were valid as liquidated damages, and we cannot deal with the error by deciding the validity issue on our own. [235 Cal. App. 3d 1396]
3. The question of prejudice.
[5a] The question remains, however, whether the error is reversible. Wells Fargo contends the error is reversible per se, while the plaintiffs argue the error is not reversible without a showing of actual prejudice.
The plaintiffs rely on decisions which state generally that " 'prejudice will not be presumed from the fact that trial was to the court or to a jury.' " (McIntosh v. Bowman (1984) 151 Cal. App. 3d 357, 363 [198 Cal. Rptr. 533], quoting Byram v. Superior Court (1977) 74 Cal. App. 3d 648, 653 [141 Cal. Rptr. 604]; see also Gann v. Williams Brothers Realty, Inc. (1991) 231 Cal. App. 3d 1698, 1704 [283 Cal. Rptr. 128]; Glogau v. Hagan (1951) 107 Cal. App. 2d 313, 318-319 [237 P.2d 329].) At first glance, these decisions would seem contrary to the well-settled rule that the denial of the constitutional right to a jury trial is reversible per se, without the need to show actual prejudice. (People v. One 1941 Chevrolet Coupe (1951) 37 Cal. 2d 283, 300 [231 P.2d 832]; Selby Constructors v. McCarthy (1979) 91 Cal. App. 3d 517, 527 [154 Cal. Rptr. 164].) Narrowly construed, however, these cases may be harmonized with the per se rule. Each involved situations where the trial court exercised its discretion to deny relief from a waiver of the right to a jury trial. (Gann v. Williams Brothers Realty, Inc., supra, 231 Cal.App.3d at p. 1704; McIntosh v. Bowman, supra, 151 Cal.App.3d at p. 363; Byram v. Superior Court, supra, 74 Cal.App.3d at p. 650; Glogau v. Hagan, supra, 107 Cal.App.2d at p. 318.) In effect, the appellate courts were reviewing the discretionary denial of relief from the waiver, not the denial of the underlying constitutional right to a jury trial. Hence, it was appropriate in those cases to invoke a traditional harmless error analysis. Any broader construction of those cases as supporting a finding of harmless error where the constitutional right to a jury trial was denied would run afoul of the reversible per se rule set forth by the California Supreme Court in People v. One 1941 Chevrolet Coupe, supra.
But none of these cases is really pertinent, as each involved the denial of a jury trial, whereas here we deal with the erroneous submission of an issue to the jury. The per se rule has no application here because it is based on the fact that the erroneous denial of a jury trial implicates an inviolate constitutional right. (See Cal. Const., art. I, § 16.) The erroneous grant of a jury trial or the improper submission of an issue to the jury is nothing more than a nonconstitutional procedural error.
Wells Fargo cites A-C Co. v. Security Pacific Nat. Bank (1985) 173 Cal. App. 3d 462 [219 Cal. Rptr. 62] for the proposition that the submission of equitable issues to a jury is reversible per se unless the trial court treated the verdict on those issues as merely advisory and made its own independent [235 Cal. App. 3d 1397] findings. But the court in A-C Co. did not engage in any reversible error analysis. The court merely noted that the trial judge had erroneously submitted an equitable issue to the jury without making independent findings, rejected the respondents' arguments that the error had been waived below, and reversed the judgment without discussing actual or presumed prejudice. (Id., at pp. 474-475.) Thus, the opinion in A-C Co. is not helpful on this point.
Absent any constitutional dimension to the present error or any existing holding that such error is reversible per se, we are left to our own devices to decide whether this nonconstitutional procedural error is reversible per se or is reversible only upon a showing of actual prejudice. This leads us inexorably to the traditional harmless error analysis for nonconstitutional error, under which prejudice is not presumed. (Code Civ. Proc., § 475; People v. Watson (1956) 46 Cal. 2d 818, 835 [299 P.2d 243].) Rather, the presumption is indulged that Wells Fargo had a fair trial, and Wells Fargo has the burden of showing otherwise. (Pool v. City of Oakland (1986) 42 Cal. 3d 1051, 1069 [232 Cal. Rptr. 528, 728 P.2d 1163].) The bank has done absolutely nothing to overcome this presumption, making no attempt whatever to show actual prejudice. Indeed, it is difficult to conceive how actual prejudice might be demonstrated in such a situation.
 This does not mean that a litigant is without any effective remedy when the trial judge erroneously decides to submit an issue to a jury. Immediate review by extraordinary writ is available, and on such pretrial review the question of prejudice does not arise. (See Gann v. Williams Brothers Realty, Inc., supra, 231 Cal.App.3d at p. 1704.) Wells Fargo had ample opportunity to make all its current jury trial arguments on immediate writ review-the trial judge having ruled on these points before trial began-but failed to do so. Indeed, the bank's counsel did not even present the trial judge with a discussion of the pre-1850 common law, evidently waiting until the appeal process before fully researching the point. Thus, Wells Fargo is itself responsible for the fact that it now faces the virtually insurmountable obstacle of traditional harmless error analysis.
The situation is analogous to several types of errors in the field of criminal law.  For example, if a nonjurisdictional irregularity in preliminary examination procedures is asserted before trial by petition for an extraordinary writ, prejudice is presumed, but on appeal after judgment the reviewing court will not reverse unless the defendant shows actual deprivation of a fair trial or other prejudice. (People v. Pompa-Ortiz (1980) 27 Cal. 3d 519, 529 [165 Cal. Rptr. 851, 612 P.2d 941].)  Similarly, prejudice is presumed on pretrial review but not on appeal where a criminal defendant is denied the statutory right to a speedy trial (People v. Wilson (1963) 60 Cal. 2d 139, [235 Cal. App. 3d 1398] 151-152 [32 Cal. Rptr. 44, 383 P.2d 452]), a change of venue mandated by extensive pretrial publicity (People v. Williams (1989) 48 Cal. 3d 1112, 1125 [259 Cal. Rptr. 473, 774 P.2d 146]), or representation by counsel of choice (People v. Chavez (1980) 26 Cal. 3d 334, 348-349 [161 Cal. Rptr. 762, 605 P.2d 401]). (See generally People v. Pompa-Ortiz, supra, 27 Cal.3d at p. 529.)
[5b] This is not to say that Wells Fargo waived the error by failing to pursue writ relief. A writ petition was not a prerequisite to appellate review. It just means that, as a practical matter, by omitting to seek immediate writ review, the bank presented itself with the daunting (perhaps impossible) task of proving actual prejudice on appeal.
This approach is fundamentally sound judicial policy. It would be a colossal waste of judicial resources to reverse a judgment after a lengthy trial, based on nonconstitutional procedural error of no demonstrated prejudice, where the error could have been cured before trial through the expediency of extraordinary writ proceedings.
We therefore conclude that although the judge erred in submitting the validity issue to the jury, the error is not reversible because Wells Fargo has failed to demonstrate any actual prejudice from the error.fn. 5
B. Monetary Relief Under Civil Code Section 1671.
 Wells Fargo next contends the plaintiffs were not entitled to monetary relief under Civil Code section 1671,fn. 6 which governs the validity of liquidated damages. The bank draws an analogy to Dean Witter Reynolds, Inc. v. Superior Court (1989) 211 Cal. App. 3d 758 [259 Cal. Rptr. 789]. The court in Dean Witter held that section 1670.5, which prescribes the defense of unconscionability to enforcement of a contract, does not create an affirmative cause of action but merely codifies the common law defense. (211 Cal.App.3d at p. 766.)
Sections 1670.5 and 1671 are indeed similar, to the extent that their enactment did not create new law but simply codified the existing common law. The statutory provisions governing liquidated damages, enacted in 1872 as sections 1670 and 1671, "announced no new principles of law in this state." (Weinreich E. Co. v. A. J. Johnston Co. (1915) 28 Cal. App. 144, 150 [151 P. 667].) The 1872 legislation provided that a contractual predetermination of damages for breach was void unless "it would be impracticable or [235 Cal. App. 3d 1399] extremely difficult to fix the actual damage." (Former § 1671.) This was a preexisting rule of common law, enunciated in California as early as 1856, when the Supreme Court stated that stipulated damages were permitted when actual damages "were wholly uncertain and incapable of estimation otherwise than by mere conjecture ...." (Cal. Nav. Co. v. Wright (1856) 6 Cal. 258, 262- 263, italics omitted; see also Nash v. Hermosilla (1858) 9 Cal. 584, 587-588.) Thus, the statutory provisions merely restated the preexisting common law.
In 1977, section 1671 was amended and former section 1670 was repealed. As amended, section 1671 prescribes a new general rule favoring liquidated damages (§ 1671, subd. (b)), but the statute retains the former codified common law rule for consumer actions. (§ 1671, subd. (d); see Cal. Law Revision Com. com., West's Ann. Civ. Code, § 1671, p. 497 ["In a consumer case, the prior law under former Sections 1670 and 1671, continued in subdivision (d), still applies."].)fn. 7
The state of the preexisting common law of liquidated damages makes the present case fundamentally different from Dean Witter. As Wells Fargo has itself pointed out on the jury trial issue, since before the enactment of the Statute of William in 1697, a party could sue in equity for affirmative relief from a contractual predetermination of damages for breach, and after 1697 relief from excessive fees could be sought at law in a defensive posture. (Ante, pp. 1391-1392.) If monetary relief could be sought under the common law of liquidated damages, and section 1671, subdivision (d), represents a codification of the theoretical underpinnings of such relief, then the codification necessarily encompasses the notion of seeking such relief. In contrast, the enactment of section 1670.5 does not allow an independent cause of action because that statute merely codifies a defense.
Wells Fargo argues that section 1671 cannot support monetary relief unless the plaintiff concomitantly asserts some other cause of action such as an accounting or money had and received. But the decisions since 1872 have been remarkably unconcerned with the label attached to a cause of action when a party seeks relief from a contractual penalty. Affirmative recovery of sums retained as purported liquidated damages has occasionally been sought on the theory of money had and received (e.g., Drew v. Pedlar (1891) 87 Cal. 443, 451-452 [25 P. 749]; Cook v. King Manor and Convalescent Hospital (1974) 40 Cal. App. 3d 782, 785 [115 Cal. Rptr. 471]; McInerney v. Mack (1917) 34 Cal. App. 153, 154 [166 P. 867]), but no decision has restricted the recovery to that theory, and courts have just as often permitted actions for [235 Cal. App. 3d 1400] affirmative relief without expressing a preference for any particular pleading theory (e.g., Freedman v. The Rector (1951) 37 Cal. 2d 16, 20-23 [230 P.2d 629, 31 A.L.R.2d 1]; Caplan v. Schroeder (1961) 56 Cal. 2d 515, 519-521 [15 Cal. Rptr. 145, 364 P.2d 321]; Clermont v. Secured Investment Corp. (1972) 25 Cal. App. 3d 766, 768-771 [102 Cal.Rptr. 340]).
Perhaps the most telling case on the unimportance of the theory of pleading is Freedman v. The Rector, supra, 37 Cal. 2d 16. There the court held that a breaching plaintiff who had sued for specific performance of a real property sales contract, while not entitled to specific performance, was entitled to recover the excess of his down payment over the damages caused by the breach, based on invalidity under the liquidated damages statute. The court directed a retrial on the latter issue, but said nothing about any necessity to amend the pleadings to add a cause of action for such recovery, even though the suit was only for specific performance. (Id., at p. 23.) Evidently, it mattered little to the Supreme Court how the action was pled; the important point, as expressed in the opinion by Justice Traynor, was that "the damage provisions of the Civil Code, together with the policy of the law against penalties and forfeitures," provided an independent "basis for relief ...." (Id., at p. 20.)
More recent decisions by the California Supreme Court indicate (without expressly holding) that affirmative relief may be sought based on section 1671, subdivision (d). In Garrett v. Coast & Southern Fed. Sav. & Loan Assn., supra, 9 Cal. 3d 731, 734- 735, the court held that the plaintiffs had stated a cause of action to recover fees paid for failure to make timely installment payments on promissory notes secured by deeds of trust, on the ground the fee assessments were void under the liquidated damages statute. The court did not expressly hold that the statute provided for affirmative monetary relief, but that was obviously the underlying assumption. Similarly, in Bondanza v. Peninsula Hospital & Medical Center (1979) 23 Cal. 3d 260, 265, footnote 6 [152 Cal. Rptr. 446, 590 P.2d 22], the court said that "injunctive relief may be predicated in the case at bar on present section 1671," again demonstrating an underlying assumption that the statute provided for affirmative relief. Even Freedman v. The Rector, supra, 37 Cal.2d at page 20, in stating that the damages provisions of the Civil Code and the policy against penalties and forfeitures provided an independent basis for relief, implies that an affirmative cause of action underlies the liquidated damages statute.
In short, this is all much ado about very little. The holding in Dean Witter has no application here because, unlike the common law as codified by section 1670.5, which provides only a defense of unconscionability, the [235 Cal. App. 3d 1401] common law as codified by section 1671, subdivision (d), permits a consumer to seek monetary relief, both offensively and defensively, from liquidated damages. We can only conclude that monetary relief may be sought under subdivision (d) of section 1671, based on (1) the preexisting common law, (2) our Supreme Court's repeated assumptions in Garrett and Bondanza that affirmative relief may be predicated on the statutory incarnation of the common law rule, and (3) the Supreme Court's striking lack of concern in Freedman for the precise pleading theory.fn. 8
C. Statute of Limitations
 Next, Wells Fargo contends the applicable statute of limitations was the two-year statute for an action on a nonwritten contract (Code Civ. Proc., § 339, subd. 1) instead of the four-year statute for an action on a written contract (id., § 337, subd. 1), and thus the judgment is flawed to the extent it covers a four-year period.fn. 9
The Supreme Court in Garrett v. Coast & Southern Fed. Sav. & Loan Assn., supra, 9 Cal.3d at page 735, commented, without expressly deciding, that "the applicable period of limitations" on that action for recovery of void liquidated damages was four years. The only possible basis for this assumption was the underlying contract requiring the payment of fees. The four-year limitation period for an action on a written contract (Code Civ. Proc., § 337, subd. 1) applies not only to the breach of an express obligation under a written contract, but also to breach of an implied promise arising out of the contract (Amen v. Merced County Title Co. (1962) 58 Cal. 2d 528, 532 [25 Cal. Rptr. 65, 375 P.2d 33]; Bruckman v. Parliament Escrow Corp. (1987) 190 Cal. App. 3d 1051, 1058 [235 Cal. Rptr. 813]; see 3 Witkin, Cal. Procedure (3d ed. 1985) Actions, § 365, pp. 393-395). Unless we are to dismiss the comment in Garrett as a careless misstatement, we must conclude that the Supreme Court viewed the retention of excessive assessments which cannot qualify as liquidated damages as the breach of an implied promise arising out of the contract.
We are not, however, restricted to such a theory here. This is because the underlying written contract-the customer agreement and disclosure statement-expressly stated in pertinent part that "this Agreement and your [235 Cal. App. 3d 1402] account, as well as our rights and duties and your rights and duties regarding this Agreement and your account, will be governed by and interpreted in accordance with the laws of the State of California ...." By expressly invoking California law, the contract impliedly invoked the rule that a void liquidated damages provision will result in liability only for actual damages. Thus, we may infer from the express provisions of the agreements, not merely the ineffectiveness of the liquidated damages provisions, an obligation to retain only the amount of actual damages. On either theory the four-year statute applies.
D. Liability for Indirect Costs and Collection Agency Fees.
We come, finally, to Wells Fargo's arguments on the merits, which concern the extent to which the plaintiffs were liable for the bank's actual damages resulting from late and overlimit activity.  The applicable rule is set forth in Garrett v. Coast & Southern Fed. Sav. & Loan Assn., supra, 9 Cal.3d at page 741, which involved the closely related subject of late charges for untimely installment payments on promissory notes secured by deeds of trust. The Supreme Court stated in Garrett that when such a charge is void, the borrower "remains liable for the actual damages resulting from his default. The lender's charges could be fairly measured by the period of time the money was wrongfully withheld plus the administrative costs reasonably related to collecting and accounting for a late payment." (Ibid.)
1. Indirect costs.
 Wells Fargo contends the Garrett measure of actual damages includes not only the "direct" increased costs to the bank resulting from specific late and overlimit activity by particular customers, but also "indirect" costs, such as overhead, that are attributable to the bank's maintenance of an "in place" infrastructure for dealing with late and overlimit activity generally.fn. 10
This contention flows inevitably from the evidence given by the bank's expert witness, Paul Regan, who testified that he made no effort whatever to determine costs resulting from particular late and overlimit activity, but instead chose to focus solely on "the kinds of costs that Wells puts in place to deal with late and overlimit customers," or, stated another way, costs incurred "as a result of there being late and overlimit conditions." Indeed, [235 Cal. App. 3d 1403] the bank concedes in its reply brief that indirect costs "were not increased by virtue of any individual late or overlimit transaction."
The problem with the bank's position is that, while Regan's approach might have been sound from an accounting perspective, it did not comport with the manner of proof required by Garrett, in particular the element of causation. Garrett authorizes assessment of damages on a theory of compensation for "actual damages resulting from [the borrower's] default" (9 Cal.3d at p. 741, italics added)-not simply, as Wells Fargo contends, for the bank's "actual losses." This standard requires a direct causal link between the breaches underlying the litigation and the actual damages caused by those breaches. (See, e.g., Guntert v. City of Stockton (1976) 55 Cal. App. 3d 131, 149 [126 Cal. Rptr. 690, 127 Cal. Rptr. 602], citing Allen v. Gardner (1954) 126 Cal. App. 2d 335, 344- 345 [272 P.2d 99] [no recovery of damages for overhead costs not increased by breach of contract].)
Regan chose to quantify the bank's damages by a method which did not make this causal connection to the underlying breaches, but instead was linked only to the general proposition that breaches may occur. The Garrett method and Regan's method do not necessarily yield the same results. For example, if there were absolutely no late or overlimit activity in a given month, Regan's method would still afford compensation for that month's costs of the "in place" infrastructure, while the Garrett method would not, since there were no underlying breaches. In view of this variance from the Garrett standard of proof, the record simply does not support Wells Fargo's theory of liability. This is not to say that indirect costs are noncompensable as a matter of law, but rather that they were not shown to be compensable here for want of any evidence of the necessary causal link.fn. 11
Even if we construe Garrett so broadly as to encompass the bank's theory of compensable indirect costs of late and overlimit "conditions," that does not mean we must accept Regan's opinion as to the amount of those costs. In fact, the amount of Wells Fargo's claimed indirect costs was hotly disputed at trial. The plaintiffs' expert, John Lehman, testified that Regan's estimate of the bank's damages was wrong not only because it did not connect costs to specific collecting or accounting activities and included overhead that would have existed even without late or overlimit activity, but also because it overstated many costs. Indeed, most of Lehman's reduction of Regan's cost estimate was attributable to the overstatement theory. [235 Cal. App. 3d 1404]
For example, Lehman testified that he eliminated Regan's inclusion of costs for "authorization"-the procedure for merchants to obtain telephone preapproval of a credit card charge-because these costs were attributable to "transactions that haven't taken place yet" and thus could not be considered costs of collecting late or overlimit balances. Lehman eliminated costs of "customer service"-the department that receives telephone inquiries from customers about their accounts-because it "doesn't make any sense" to assess costs for incoming calls. He eliminated costs of "cardholder accounting"-a department which performs general ledger accounting and file maintenance such as name and address changes-because "their work effort wouldn't vary at all based on the number of late or overlimit accounts." He eliminated some costs of "delinquency control" for functions unrelated to late and overlimit activity, such as public relations, credit card fraud, billing disputes and the removal of stolen cards from circulation. He eliminated "accounts research" costs pertaining to customers' claims that they did not make charges appearing on their bills.fn. 12
In each of these instances, Lehman's testimony constituted an expression of opinion that under the Garrett standard the disputed costs were not "reasonably related" to collecting and accounting for late or overlimit balances because those costs had nothing to do with any late or overlimit activity, not because they were indirect costs not linked to particular late and overlimit transactions. Whether the disputed indirect costs "resulted" from late or overlimit activity or, as Lehman asserted, from other credit card functions, was a classic question of fact for a jury to decide.
Thus, the indirect cost issue reveals itself to be a "question of fact" masquerading in "question of law" clothing. There was an evidentiary conflict as to whether Regan had overstated the amount of Wells Fargo's indirect costs. On appeal, pursuant to the substantial evidence rule, we are not free to reject Lehman's opinion in favor of Regan's estimate, but must resolve the evidentiary conflict in favor of the judgment. (E.g., Hasson v. Ford Motor Co. (1977) 19 Cal. 3d 530, 544 [138 Cal. Rptr. 705, 564 P.2d 857, 99 A.L.R.3d 158].) The jury, by returning a verdict for approximately half the amount sought by the plaintiffs, can be said to have, in effect, accepted half of Lehman's reductions. Lehman's adjustments for authorization, customer service, cardholder accounting, delinquency control and accounts research in themselves accounted for about half of his total reductions. The full amount of reductions for overstated costs (excluding overhead) greatly exceeded half of Regan's cost estimate. Accordingly, even accepting Wells Fargo's theory of indirect costs recovery, the judgment is unaffected because [235 Cal. App. 3d 1405] substantial evidence supports the jury's apparent rejection of half the claimed amount of indirect costs.
2. Collection agency fees.
 Finally, Wells Fargo contends the Garrett standard of compensation for actual damages includes fees the bank paid to collection agencies in its attempts to collect late (but not overlimit) balances. The bank characterizes these fees, which are a percentage (commonly one-third) of a collected debt, as "direct" costs.
This issue is resolved by Bondanza v. Peninsula Hospital & Medical Center, supra, 23 Cal. 3d 260. In that case, the Supreme Court held that a hospital's assessment of collection agency fees against patients, based on their contractual promise to pay a "reasonable collection expense," was impermissible under Garrett. Focusing on the restriction of compensation to actual damages when a liquidated damages provision is void, the court explained that "there may be no relationship whatever between the charge assessed against the patient and the actual expense required to collect an account." (Id., at p. 267, italics added.) For example, the obligation might be mostly or fully paid shortly after assignment to the collection agency "and before the collection agency has expended any significant effort to collect the debt." (Ibid.) Without this necessary relationship to the actual cost of collection, the fees were not compensable, even though the hospital had actually paid them. (Id., at p. 268.) The court rejected the notion that the fees were justified by the need to reimburse the collection agency for expenses incurred in pursuing uncollectible accounts, concluding that the patients whose accounts were collected "cannot lawfully be required to guarantee the economic well-being of the agency." (Id., at pp. 268-269.)
The reasoning of Bondanza is dispositive here. If, to paraphrase the Bondanza opinion, the amount of a collection agency fee may have no relationship to the actual expense required to collect a late balance, the fee does not necessarily represent Wells Fargo's "actual damages resulting from [the borrower's] default" within the meaning of the Garrett standard (9 Cal.3d at p. 741), even though the bank actually paid the fees. Only the actual expenses of collection (which the bank did not demonstrate) were compensable.
Wells Fargo claims Bondanza is distinguishable because compensation for its collection agency fees would spread the responsibility for those fees among all its late debtors, and thus avoid the problem of debtors with collected accounts bearing the full cost of pursuing uncollected accounts. This argument is flawed in that it depends on an unwarranted assumption [235 Cal. App. 3d 1406] that Wells Fargo's collection agencies had no other clients. Unless we indulge that assumption, the problem described in Bondanza still exists, since the bank's debtors would still bear each collection agency's cost of pursuing the uncollected accounts of its other clients.
In any event, even if Wells Fargo's collection agency fees were compensable, the judgment remains unaffected, since the verdict is supported by Lehman's reductions for overstated indirect costs independent of his exclusion of collection agency fees.fn. 13
The judgment is affirmed.
Perley, Acting P. J., and Stein, J.,fn. * concurred.
FN *. Associate Justice of the Court of Appeal, First District, Division One, sitting under assignment by the Chairperson of the Judicial Council.
FN 1. There is more at stake in this appeal than the $5,227,617 judgment. Wells Fargo settled a similar action against Crocker National Bank (which was acquired by Wells Fargo in 1986) for $3,782,273, but the settlement is conditioned on the outcome of this appeal. Pursuant to the plaintiffs' request, we take judicial notice of the settlement agreement in that case. (Kovitz v. Crocker National Bank (Super. Ct. S.F. City & County, 1990, No. 868914).)
FN 2. Because of this conclusion, we need not reach Wells Fargo's argument that there was no right to a jury trial because the plaintiffs' suit was essentially an equitable action for an accounting.
FN 3. We therefore need not address the plaintiffs' contention that the bank's objection below was untimely.
FN 4. The plaintiffs rely on Atkinson v. Pacific Fire Extinguisher Co. (1953) 40 Cal. 2d 192, 195 [253 P.2d 18], which said that the validity question is one of fact, but the court in that case did not address or decide whether the judge or jury decides the question. (See post, p. 1394.) Wells Fargo relies on several pre-1850 English cases in which trial judges decided validity questions, but, again, those opinions did not expressly hold that this was a matter for the judge to decide. Counsel have evidently overlooked Sainter v. Ferguson, supra, in which the applicable rule is expressly stated.
FN 5. Even if error were found as to the first issue raised by the bank (see pt. III.A.1., ante), we would conclude, by this same reasoning, that the error would not be reversible.
FN 6. All further statutory references are to the Civil Code unless otherwise indicated.
FN 7. Wells Fargo has expressly declined to challenge the application of section 1671, subdivision (d) to this case, effectively conceding that this is a "consumer" action.
FN 8. The question whether section 1671 permits monetary relief was recently presented to Division Two of this court in a writ petition, which the court summarily denied. (McClendon v. Superior Court (Aug. 1, 1990) A050043 [nonpub. opn.].) Pursuant to Wells Fargo's request, we take judicial notice of the pleadings and order in that proceeding. The point is meaningless, however, since the summary denial has no precedential effect. (E.g., People v. Medina (1972) 6 Cal. 3d 484, 490-491 [99 Cal. Rptr. 630, 492 P.2d 686].)
FN 9. The plaintiffs claim that Wells Fargo waived the statute of limitations argument by failing to assert it below. This is untrue; the bank made the argument in its trial brief.
FN 10. Under Wells Fargo's theory, all such costs are encompassed by this action because the class aspect of the case implicates all late and overlimit activity.
FN 11. Wells Fargo draws analogies to other fields of law and taxation where indirect costs such as overhead are treated like direct costs, but the direct applicability of the standard announced in Garrett makes any variance from that standard by analogy to such other fields unnecessary and inappropriate.
FN 12. Lehman also testified that certain other costs "would exist anyway, regardless of whether people were over limit or late," but this testimony pertained only to overhead costs.
FN 13. Wells Fargo also asserts that Lehman's testimony should have been excluded or stricken, and that the court erred in rejecting the bank's proposed jury instructions on indirect costs and omitting to instruct the jury on the compensability of collection agency fees. These assertions, however, depend wholly on Wells Fargo's claims that Lehman's testimony was inconsistent with the theory that the bank's indirect costs and collection agency fees were recoverable as a matter of law. Those arguments having failed, the attacks on Lehman's testimony and the absence of jury instructions necessarily fail.
FN *. Associate Justice of the Court of Appeal, First District, Division One, sitting under assignment by the Chairperson of the Judicial Council.