People v. Ashford University CA4/1

Filed 2/20/24 P. v. Ashford University CA4/1 NOT TO BE PUBLISHED IN OFFICIAL REPORTS California Rules of Court, rule 8.1115(a), prohibits courts and parties from citing or relying on opinions not certified for publication or ordered published, except as specified by rule 8.1115(b). This opinion has not been certified for publication or ordered published for purposes of rule 8.1115. COURT OF APPEAL, FOURTH APPELLATE DISTRICT DIVISION ONE STATE OF CALIFORNIA THE PEOPLE OF THE STATE OF D080671 CALIFORNIA, Plaintiff and Respondent, (Super. Ct. No. 37-2018- v. 00046134-CU-MC-CTL) ASHFORD UNIVERSITY, LLC, et al., Defendants and Appellants. APPEAL from a judgment of the Superior Court of San Diego County, Eddie C. Sturgeon, Judge. Affirmed as modified. Gibson, Dunn & Crutcher, Theane D. Evangelis, Jeremy S. Smith, Daniel M. Rubin and Ryan Azad for Defendants and Appellants. Rob Bonta, Attorney General, Nicklas A. Akers, Assistant Attorney General, Michael E. Elisofon, Emily C. Kalanithi, Rachel A. Foodman, Joseph M. Lake and Bree B. Baccaglini, Deputy Attorneys General, for Plaintiff and Respondent. INTRODUCTION Zovio, Inc. (Zovio) and Ashford University, LLC, a California limited liability company (together with Zovio, defendants) are the former owners and operators of an online university. Following a bench trial, the trial court found that for more than a decade, defendants violated the unfair competition law (UCL) (Bus. & Prof. Code,1 § 17200 et seq.) and false advertising law (FAL) (§ 17500 et seq.) by making false and misleading statements to prospective students. The court determined defendants committed 1,243,099 UCL and FAL violations and imposed $22,375,782 in civil penalties. Defendants do not challenge the trial court’s liability determination. Instead, their appeal is focused on obtaining a reduction in the $22,375,782 penalty award. They make the following contentions: (1) the penalty should be reduced because the court included in the penalty count violations occurring outside the FAL’s statute of limitations; (2) the court erred in the manner in which it calculated the number of violations; (3) the penalty should be reduced to avoid violating extraterritoriality principles; (4) the penalty should be reduced because it does not bear a reasonable relationship to the harm proven at trial; and (5) the penalty should be reduced because it is excessive given defendants’ financial condition. We agree with defendants’ first contention and conclude the trial court inadvertently penalized defendants for expired FAL violations. Although defendants ask us to reduce the penalty by $939,279, we conclude this overstates the extent of the error. We shall reduce the judgment by $933,453. 1 Further undesignated statutory references are to the Business and Professions Code. 2 We are unpersuaded by defendants’ remaining claims of error. Accordingly, we affirm the judgment as modified. FACTUAL AND PROCEDURAL BACKGROUND I. Ashford University2 Zovio is an education technology services company that provides student enrollment services to higher education institutions. Until December 2020, Zovio, through its wholly-owned subsidiary, Ashford University, LLC, also owned and operated Ashford University. Ashford University was founded in 2005, when Zovio, which had never offered any degree programs, purchased a small campus-based religious university located in Iowa. Zovio needed this university’s accreditation because only students attending accredited institutions are eligible for federal financial aid. Zovio renamed the school Ashford University (Ashford) and transformed it into an enormous online institution that was marketed as a traditional university. In December 2020, a California entity affiliated with the University of Arizona acquired Ashford and rebranded it as the University of Arizona Global Campus (UAGC). A typical bachelor’s degree from Ashford cost between $40,000 and $60,000. At its peak, Ashford had more than 80,000 students and generated hundreds of millions of dollars annually, the vast majority from taxpayer- 2 Our factual summary is derived principally from the trial court’s statement of decision, the accuracy of which has not been challenged in this appeal. (See Rael v. Davis (2008) 166 Cal.App.4th 1608, 1617 [where defendants do not contend the trial court’s factual findings are unsupported by substantial evidence, the facts set forth in the statement of decision are accepted as true].) 3 funded sources such as Title IV loans, income-based grants, and G.I. Bill funds. The students who enrolled at Ashford were described by Ashford’s former presidents as leading “complex” and “difficult lives.” Based on Zovio’s own assessments from 2009 through 2020, Ashford students were typically older than traditional college students with an average age of 35 to 37, and of low income (between 55 and 76 percent received Pell Grants3). Around half of Ashford students identified as minorities. Only a quarter of Ashford students graduated, and many defaulted on their student loans. Former presidents of Ashford recognized that the vulnerability of its student population “heighten[ed]” the need for accurate college advising. Even so, defendants created a high-pressure admissions department whose “north star” was enrollment numbers rather than truthful advising. Defendants enrolled students in Ashford primarily through the use of sales people—whom Ashford referred to as “admissions counselors”—trained to build trust and rapport with prospective students. Admissions counselors were expected to call hundreds of leads every day. Managers would threaten to fire those who failed to enroll enough students in Ashford, warning that “ ‘[s]omeone can fill your chair’ ” if counselors did not meet their numbers. Defendants’ admissions counselors described the work environment as permeated by fear and as a place where “closing the sale” was prioritized above providing prospective students with accurate information. One of Ashford’s former presidents received emails warning that fear was “abundant” in the admissions department and 3 A Pell Grant is a subsidy the U.S. federal government provides through participating institutions for students with exceptional financial need who have not earned their first bachelor's degree or who are enrolled in certain post-baccalaureate programs. 4 enrollment numbers were seen as the “end-all-be-all.” Defendants’ own employee exit surveys further confirmed that employees experienced the admissions department as a “boiler room” driven by fear of not enrolling enough students. In this environment, admissions counselors succumbed to the pressure and made deceptive statements to prospective students in order to boost their enrollment numbers and keep their jobs. II. The Enforcement Action In November 2017, the Attorney General filed an enforcement action against defendants on behalf of the People of the State of California. In the operative complaint, the People alleged defendants had violated the UCL and FAL by making myriad misrepresentations to prospective Ashford students regarding the costs of attending Ashford, the availability of financial aid, the ability of Ashford programs to prepare students for careers in certain professions, and the likelihood that academic credits would transfer into and out of Ashford. They also alleged that defendants had employed unfair, unlawful, and fraudulent billing and collections practices in violation of the FAL and UCL. They sought injunctive relief, restitution, and civil penalties of $2,500 for each UCL violation and $2,500 for each FAL violation. Effective February 6, 2013, the parties executed an agreement tolling all time-related defenses, including defenses based on the statute of limitations. The UCL has a four-year statute of limitations (Bus. & Prof. Code, § 17208), and the FAL has a three-year statute of limitations (Code Civ. Proc., § 338, subd. (h)). Thus, the earliest that defendants could be held liable for UCL violations was February 6, 2009, and the earliest they could be held liable for FAL violations was February 6, 2010. 5 III. Trial and Statement of Decision From November 2021 until December 2021, the trial court held a bench trial during which it heard in-court testimony of 23 witnesses, considered designated deposition testimony of another 17 witnesses, and admitted 1,514 exhibits in evidence. On March 3, 2022, after receiving proposed statements of decision from both sides, the trial court issued an extremely detailed, 49-page statement of decision finding defendants liable for making over one million misleading calls in violation of the UCL and FAL. The trial court found defendants created a high-pressure, fear-based culture in the Ashford admissions department—a fact known to defendants’ executives— which prioritized enrollment numbers over providing accurate information. As a result, admissions counselors made misrepresentations to students on matters critical to students’ decision-making. The court explained that each category of misrepresentation was supported by four types of evidence: (1) the testimony of nine student victims who experienced the misrepresentations and relied on them in deciding to enroll at Ashford; (2) the testimony of former Ashford employees who explained how the pressure to meet enrollment numbers, instructions from managers, and other guidance led them to deceive students in order to promote enrollment; (3) the testimony of Dr. Jerome Lucido, an expert in college admissions with over 40 years of experience setting industry standards for college advising and leading the admissions, financial aid, and registrar departments of four major universities; and (4) internal company documents and testimony of witnesses affiliated with defendants. 6 A. The Misrepresentations The trial court described the topics on which admissions counselors made misrepresentations as well as the associated supporting evidence in great detail. There were nine such topics. 1. Falsely promising prospective students that Ashford degrees qualified students for teaching careers. First, the trial court found defendants falsely promised prospective students they could use an Ashford degree to become a teacher. In fact, Ashford degrees did not qualify graduates for teaching positions that require licensure. Students deceived into enrolling at Ashford had to invest significant additional time and money in a state-approved teaching program. For example, Alison T. testified she enrolled at Ashford because her admissions counselor assured her Ashford was part of an “interstate agreement” pursuant to which her degree would “carry over” to Pennsylvania so long as she completed her student teaching and passed state teaching exams. Only after graduating did she discover this was not the case, and she would need to complete a substantial number of additional credits before she could begin student teaching. Crystal E. testified she was misled into enrolling at Ashford, and withdrawing from a degree program at a different school that would have led to teacher licensure, because she was told Ashford’s program was equivalent to the one she was attending. Only after graduating did she discover this was not the case and her Ashford education did not qualify her to take the state teaching exam. The trial court found the testimony of these witnesses was corroborated by Dr. Lucido’s call analysis (summarized below), which identified 10 calls with at least one “teaching misrepresentation.” It further found that had defendants not misled prospective students to believe their Ashford degrees 7 would lead to licensure, they could have attended a school with a “two-in-one” four-year bachelor’s degree program that is also state-approved for teaching. 2. Misleading prospective students that Ashford degrees qualified them for “helping careers.” Second, defendants routinely misled students about their ability to use an Ashford degree to become nurses, social workers, and drug and alcohol counselors (“helping careers”), professions that also require state licensure or certification. Although Ashford degrees are not state-approved for any of these helping careers, defendants repeatedly deceived prospective students by telling them Ashford was “perfect” or “geared for” students with such aspirations. For example, Roberta P. testified her admissions counselor told her a master’s degree in psychology would allow her to work in occupations like counseling and therapy, which caused her to reasonably believe her Ashford degree would meet the requirements for a therapy license. Only after graduating with $40,000 in student loans did she discover Ashford’s program was not state-approved and she would have to complete a separate program in order to achieve her career goals. Similarly, Pamela R.’s admissions counselor told her she would have “no problem” becoming a certified substance abuse counselor with an Ashford degree. She learned one week before graduation this was not the case, and her degree did not meet any of the necessary requirements. Dr. Lucido identified seven calls with similar misrepresentations directed at the helping careers. 3. Misleading prospective students about financial aid. Third, defendants misled students about the amount of financial aid they would receive and the costs that aid would cover. Dr. Lucido identified 46 calls with misrepresentations in this category. 8 The trial court summarized student and former employee testimony confirming such statements were likely to deceive. Loren E. testified her admissions counselor promised that financial aid would cover the costs of her degree. She discovered this promise was false when she reached her lifetime loan limit before graduating and was forced to drop out, leaving her with “massive debt but no degree.” Jasmine C. (who was also told, falsely, that her Ashford degree would allow her to be a nurse) described a similar experience. Other witnesses testified that Ashford’s admissions counselors made false representations about the likely availability of Pell Grants. The trial court found that making unsupported representations about financial aid and out-of-pocket costs was misleading. Seventy-five percent of students who ultimately received financial aid did not receive their financial aid award letter until after enrollment; one-third of students did not receive their award letter until after the point when they became liable to pay tuition. The court concluded that statements in this category were deceptive and the evidence demonstrated defendants “plainly recognized that it was misleading for admissions counselors to predict aid awards or out-of-pocket costs.” 4. Downplaying future debt. Fourth, admissions counselors misled students by downplaying future debt. For example, counselors deceptively quoted students’ future loan payments at a small fraction of their potential size. Dr. Lucido identified four calls in this category, and testified they were misleading because admissions counselors cannot know what a student’s loan payments will be or how much debt a student will take on. 9 5. Misstating federal financial aid rules. Fifth, defendants misled students about the rules and requirements governing federal financial aid despite knowing it was deceptive to misstate financial aid rules. Dr. Lucido identified eight calls with such misrepresentations. 6. Misleading prospective students on the feasibility of “doubling up” on class credits. Sixth, defendants misled students about the feasibility of taking two classes simultaneously (or “doubling up”) rather than one class at a time, which was the standard. Doubling up can generate significant out-of-pocket costs due to financial aid limits. Defendants knew it was misleading to tell students about doubling up on classes without also disclosing the additional costs. Even so, 30 calls were identified by Dr. Lucido as containing this type of misrepresentation. 7. Understating costs of attendance. Seventh, defendants misled students about the cost of an Ashford degree in several ways: by leading students to believe their only cost would be tuition, when in fact they would incur significant additional expenses for books and fees; by quoting costs that did not match the academic catalog (e.g., understating the cost of a degree program); by inaccurately comparing Ashford’s price with that of other schools (e.g., by saying U.C. Berkeley is more expensive than Ashford, when Ashford costs more for the same number of credits); and by quoting students the cost of an academic year, which was misleading because an Ashford bachelor’s degree is divided into five “academic years” rather than the traditional four years. Former student Alison T., for example, testified an admissions counselor told her the cost of Ashford was around $10,000 per academic year, leading her to reasonably believe her degree would cost around $40,000, 10 when in fact it cost more than $50,000. Dr. Lucido identified a total of 33 calls with misrepresentations in this category. 8. Misrepresenting the pace and time commitment for completion. Eighth, defendants misrepresented the pace and time commitment for completing an Ashford degree by mischaracterizing its bachelor’s degree program as akin to traditional four-year programs. In fact, whereas students at traditional schools earn 30 credits between September and May, allowing them to finish a 120-credit degree in four years, a typical Ashford student must take classes for 50 weeks per year (with no summer break) in order to earn the same 30 credits. Dr. Lucido identified a total of 29 calls in which admissions counselors made these sorts of misrepresentations. 9. Misleading prospective students on the ability to transfer credits. Ninth, defendants misled students about the ability to transfer credits in and out of Ashford. Transfer credits matter because they can reduce the time and cost of a degree. And yet admissions counselors “routinely” made inaccurate promises that students’ prior credits, or life experience, would transfer. For example, Jessica O.’s admissions counselor told her at least half of her prior credits would transfer into Ashford “no matter what.” Only after completing her first class did she learn less than one-third of her prior credits had transferred, extending the time required to complete her degree. The trial court found the testimony of students like Jessica “mirror[ed]” 39 calls identified by Dr. Lucido as containing at least one misrepresentation about students’ ability to transfer credits to Ashford. Lucido identified four additional calls in which defendants gave students misleading assurance their Ashford credits would transfer elsewhere. Although defendants knew it was misleading to promise or imply credits would transfer, former Ashford 11 employees testified that misrepresentations “like the ones Dr. Lucido identified” were routinely made and even encouraged by managers. B. Defendants’ Knowledge of Deception Next, the trial court found defendants were “well aware of the deception pervading their admissions department,” and had “amassed an extensive paper trail documenting the same misrepresentations identified by Dr. Lucido.” Greg Regan, the People’s forensic accounting expert, analyzed defendants’ scorecard data (documents defendants’ compliance department used to assess calls between employees and students) and determined admissions counselors made relevant non-compliant statements in 20.5 percent of scorecards. A compliance director responsible for admissions call monitoring observed in a 2010 email there were “areas where the level of negligence is astonishing.” And from 2012 to 2014, defendants received “mystery shopper reports” documenting specific misrepresentations about financial aid and transfer credits. Although these reports revealed systematic deception in admissions, defendants did not take the findings seriously. Instead, they discontinued the mystery shopper program. Finally, executives received “troubling complaints” directly from students but failed to take appropriate action. The trial court further found that defendants tended to promote, rather than meaningfully discipline, repeat offenders, illustrating their “knowledge and acceptance, even approval, of misrepresentations.” Defendants promoted substantial numbers of admissions counselors who made relevant non- compliant statements in at least half of their monitored calls. These promotion decisions encouraged noncompliance in line-level admissions employees. 12 C. Calculation of UCL and FAL Violations After determining that defendants were liable under the UCL and FAL for admissions counselors’ misrepresentations because defendants had the right to control their activities, and after rejecting defenses not at issue here, the trial court turned to the question of civil penalties. Citing People v. Morse (1993) 21 Cal.App.4th 259 (Morse), the court found it appropriate to count each deceptive telephone call made by defendants as a separate violation. To quantify the number of deceptive calls, the trial court relied on the expert testimony of Dr. Lucido and Dr. Bernard Siskin, a statistician, both of whom the court found to be credible. 1. California Calls 2009–2012 Between 2013 and 2020, there were 1,573,400 phone calls between defendants and California students. To determine the number of deceptive calls within this population, Dr. Siskin drew a random sample of 2,234 calls. Siskin determined there were 561 admissions calls discussing at least one relevant topic within this sample. Next, the 561 relevant calls were sent to Dr. Lucido, who reviewed them in context, highlighted any misrepresentations and assigned each misrepresentation a category code, and determined 126 (22 percent) of the relevant calls contained at least one misrepresentation. From these results, Siskin determined the total number of misleading calls to California students during this period was 88,742. Defendants did not produce phone calls for the 46-month period spanning from March 2009 through December 2012. Dr. Siskin therefore assumed that during this period, defendants made calls to California students at the same average monthly volumes, and with the same percentage of misrepresentations, reflected in his random sample. Based on Siskin’s calculations, the trial court concluded defendants made 46,386 13 misleading calls to California students from March 2009 through December 2012. 2. Total Misleading Calls Nationwide 2009–2012 Defendants retained and produced only California calls. However, relying on testimony of Dr. Siskin, the trial court determined the number of misleading calls nationwide and set forth the relevant calculations in an appendix attached to the statement of decision. Based on these calculations, the court concluded the total number of misleading calls placed by defendants from March 2009 through April 2020 was 1,243,099. D. Civil Penalties To decide the appropriate amount of civil penalties, the trial court considered the statutory penalty factors listed in the UCL and FAL (see §§ 17206, subd. (b), 17536 [“the nature and seriousness of the misconduct, the number of violations, the persistence of the misconduct, the length of time over which the misconduct occurred, the willfulness of the defendant’s misconduct, and the defendant’s assets, liabilities, and net worth”]) and determined that $22,375,782 in civil penalties was reasonable and supported by the evidence. The court declined to order an injunction or award restitution. It also ruled the People had not proven their separate claim that Zovio engaged in illegal debt collection practices. IV. Defendants’ Motion for New Trial In April 2022, defendants filed a motion for new trial or to vacate judgment on the ground the civil penalty award was excessive and unsustainable under the statutory UCL and FAL factors, violated the state and federal constitutional prohibitions on excessive fines, and was affected by an assortment of errors. After a hearing, the trial court issued a minute 14 order denying the motion on May 13. Defendants filed a timely notice of appeal.4 DISCUSSION I. Defendants’ Briefing Violations Before we analyze the merits of defendants’ challenges, we first must address the factual background section of defendants’ opening brief, which violates several principles governing the content of appellate briefs. A. Defendants’ Factual Recital Is Slanted and Unduly Argumentative The first flaw relates to the manner in which defendants have presented the relevant facts. Defendants’ factual recital is a one-sided narrative that highlights favorable testimony while ignoring or downplaying the trial court’s adverse factual findings. Defendants state, for example, that Ashford “sought to be a ‘place of opportunity’ ” for disadvantaged students, downplaying that the court found Ashford deceived those same students. They emphasize their executives’ testimony that admissions counselors’ role was to “ ‘help’ ” and “ ‘educate,’ ” ignoring that the court found defendants’ admissions counselors were sales people who were pressured to 4 Defendants’ notice of appeal identified the statement of decision and minute order denying their new trial motion as the appealed orders. “ ‘The general rule is that a statement or memorandum of decision is not appealable.’ ” (Morgan v. Imperial Irrigation Dist. (2014) 223 Cal.App.4th 892, 904.) However, reviewing courts have discretion to treat a statement of decision as appealable when it “ ‘is signed and filed and does, in fact, constitute the court’s final decision on the merits.’ ” (Ibid.) The trial court’s statement of decision has all of these attributes, so we will exercise our discretion to treat it as a final, appealable judgment. An order denying a motion for new trial is not independently appealable, but may be reviewed on appeal from the underlying judgment. (Walker v. Los Angeles County Metropolitan Transportation Authority (2005) 35 Cal.4th 15, 18.) 15 persuade prospective students to enroll. They seek to minimize the consequences of their illegal actions by pointing to what they seem to regard as favorable details from student victims’ testimony, such as that one victim is not making payments toward her Ashford student loan debt (which they appear to view as relevant to the scope of her economic losses). California Rules of Court, rule 8.204(a)(2)(C) provides that an appellant’s opening brief shall “[p]rovide a summary of the significant facts.” And the leading California appellate practice guide gives this instruction: “Before addressing the legal issues, your brief should accurately and fairly state the critical facts (including the evidence), free of bias; and likewise as to the applicable law.” (Eisenberg et al., Cal. Practice Guide: Civil Appeals and Writs (The Rutter Group 2023) ¶ 9:27, p. 9-8, italics added; accord In re Marriage of Davenport (2011) 194 Cal.App.4th 1507, 1531; Hjelm v. Prometheus Real Estate Group, Inc. (2016) 3 Cal.App.5th 1155, 1166 (Hjelm).) In some procedural scenarios, appellate courts will view the facts and evidence in the light most favorable to the appellant. (See Eisenberg et al., Cal. Practice Guide: Civil Appeals and Writs (The Rutter Group 2023) ¶ 9:145, p. 9-43 [listing judgment of dismissal after sustention of demurrer, judgment on the pleadings, summary judgment, judgment NOV or nonsuit as scenarios in which reviewing court views allegations or evidence in light most favorable to appellant].) This is not one of them. Defendants have challenged the penalty as a misapplication of law to underlying facts. They have not disputed the correctness of the factual findings in the statement of decision. Although our review of the trial court’s application of law to fact is independent, we defer to the court’s underlying factual findings and “accept as true all evidence and all reasonable inferences from the evidence tending 16 to establish the correctness of the trial court’s findings and decision[.]” (Adhav v. Midway Rent A Car, Inc. (2019) 37 Cal.App.5th 954, 969 (Adhav); see PR/JSM Rivara LLC v. Community Redevelopment Agency (2009) 180 Cal.App.4th 1475, 1486 (PR/JSM) [trial court’s unchallenged factual findings presumed correct on appeal].) Defendants’ factual narrative ignores these principles by presenting the facts and evidence in the light most favorable to themselves. Citing Nwosu v. Uba (2004) 122 Cal.App.4th 1229, 1246, defendants contend that because they have not challenged the existence of substantial evidence to support the trial court’s factual findings, they have appropriately “limit[ed] [their] discussion to the evidence necessary to understand the case and the legal issues raised on appeal.” We disagree. Defendants have not set forth all “evidence necessary to understand the case”; they have “ ‘reargue[d] the “facts” as [they] would have them[.]’ ” (Hjelm, supra, 3 Cal.App.5th at p. 1166.) And Nwosu, which involved an appellant’s violation of the briefing standards that apply on substantial evidence review, does not give defendants free reign to use their factual recital to present a slanted narrative simply because their appeal involves a different standard of review. We disapprove of the distorted narrative defendants have presented here. And while defendants deny that they have raised a sufficiency of the evidence challenge to the trial court’s factual findings, we observe that any such challenge has also been forfeited due to their briefing violation. (See Rayii v. Gatica (2013) 218 Cal.App.4th 1402, 1408 [factual presentation that cites and discusses only evidence in appellant’s favor while ignoring facts and evidence favoring respondent “waives the contention that the evidence is insufficient to support the judgment”].) 17 B. Factual Assertions Based on Matters Outside Scope of Appellate Review The next set of briefing deficiencies have to do with defendants’ factual assertions, and more specifically with their failure to properly ground those assertions on matters of record. First, defendants’ brief is peppered with factual statements, such as that Zovio is now dissolved, or that it suffered the “financial death penalty.” Although conveyed as if they are facts, these statements are not followed by a citation to the record or any supporting evidence. We disregard all such unsourced factual assertions. (See Cal. Rules of Court, rule 8.204(a)(1)(C); Fierro v. Landry’s Restaurant Inc. (2019) 32 Cal.App.5th 276, 281, fn. 5 [“appellate courts may ‘ “disregard any factual contention not supported by a proper citation to the record” ’ ”].) Second, some of defendants’ statements of fact (such as that Zovio’s financial position “worsened” after trial) are based on exhibits submitted in support of a post-trial motion they filed to obtain relief from the appeal bond requirement. “ ‘ “It has long been the general rule and understanding that ‘an appeal reviews the correctness of [an order] as of the time of its rendition, upon a record of matters which were before the trial court for its consideration.’ ” ’ ” (Glassman v. Safeco Ins. Co. of America (2023) 90 Cal.App.5th 1281, 1307 (Glassman).) The cited exhibits were not before the court when it awarded the civil penalties or when it ruled on the merits of defendants’ new trial motion in which they challenged the amount of the penalties. “It would undermine basic principles of appellate review to include within the scope of our review matters” that were not “offer[ed] to the trial court in the first instance in making the ruling now challenged on appeal.” (Ibid.) As a result, we disregard these exhibits as well as the facts derived from them. 18 Third, defendants’ opening brief includes a narrative of events intended to convey the financial “[r]uin” (boldface omitted) Zovio purportedly suffered as a result of the judgment. All of the facts in this section are derived from materials outside the scope of our review. Some are taken from exhibits filed in support of defendants’ bond motion and are extraneous for the reasons just discussed. (Glassman, supra, 90 Cal.App.5th at p. 1307.) Most are based on SEC filings published on the internet. As support, Zovio cites web addresses where the filings can be accessed directly. By referring us to such internet publications, defendants violate the rule against citing matters outside the record. (See Christ v. Schwartz (2016) 2 Cal.App.5th 440, 450, fn. 5 [“publications that are not a part of the trial record cannot be considered on appeal”].) We disregard defendants’ citations to materials published on the internet as well as the facts they purportedly contain.5 5 In a footnote, defendants ask us to take judicial notice of an SEC filing (Zovio’s 2021 10-K form) and the facts within it pursuant to Evidence Code section 452, subdivision (h). Rather than submit the document itself, they have provided us a web address where it can be accessed. We deny the request for judicial notice for the following reasons. Defendants’ request does not comply with rule 8.252(a) of the California Rules of Court, which states that in order to obtain judicial notice by a reviewing court “a party must serve and file a separate motion with a proposed order” and a copy of the matter to be judicially noticed. (Kao v. Joy Holiday (2020) 58 Cal.App.5th 199, 204, fn. 3 [denying request for judicial notice for failure to comply with this rule].) Further, “[a]n appellate court may properly decline to take judicial notice under Evidence Code sections 452 and 459 of a matter which should have been presented to the trial court for its consideration in the first instance” (Brosterhous v. State Bar (1995) 12 Cal.4th 315, 325–326) and “ ‘may decline to take judicial notice of matters not relevant to dispositive issues on appeal’ ” (Meridian Financial Services, Inc. v. Phan (2021) 67 Cal.App.5th 657, 687, fn. 10). Defendants do not establish the referenced SEC filing was provided to the trial court in the first instance, nor have they 19 In their reply brief, defendants try to validate their reliance on such extraneous matters by claiming post-trial evidence may be considered when reviewing the constitutionality of a penalty. This claim, which is based on Nickerson v. Stonebridge Life Insurance Co. (2016) 63 Cal.4th 363, 376–377, lacks merit. In Nickerson, pursuant to a stipulation of the parties, the trial court issued a postverdict award of attorney fees that represented an element of the plaintiff’s damages pursuant to a rule established in Brandt v. Superior Court (1985) 37 Cal.3d 813. (Nickerson, at pp. 369–370.) The question was whether these Brandt fees could be added to the jury’s compensatory damages award when calculating the ratio of punitive damages to compensatory damages. (Id. at pp. 372–373.) Our high court held they could, reasoning “it is not clear why judicial review should be more constrained where, as here, the parties have agreed to submit an element of the plaintiff’s harm to the trial court for resolution.” (Id. at p. 376.) As this description of Nickerson makes clear, its holding has no application to the circumstances before us. No postverdict award was issued in this case, and the usual rules of procedure were not altered by stipulation. Accordingly, Nickerson does not allow defendants to subvert the rule that a reviewing court considers only the evidence that was before the trial court when it made the rulings challenged on appeal. And while defendants try to persuade us to consider their new factual assertions by claiming they are “undisputed,” we disagree with this characterization. No stipulation has been offered to establish the truth of any of the facts they derive from materials outside the record. Moreover, the People characterize defendants’ facts as “untested” and “unsupported” and tendered an argument affirmatively demonstrating that the information they derive from it is relevant to dispositive issues on appeal. 20 ask us to disregard them. This is sufficient to convey that the new matters are not accepted as true by the People and they are, therefore, disputed. (See also fn. 5, post.) Nor do defendants show that any of the circumstances that sometimes support consideration of postjudgment developments exist here. (See Reserve Insurance Co. v. Pisciotta (1982) 30 Cal.3d 800, 813 [observing that courts of review consider postjudgment events “when legislative changes have occurred subsequent to a judgment,” “when subsequent events have caused issues to become moot,” or to avoid repetitive litigation of issues].) Therefore, we rely solely upon the evidence that was before the trial court when it rendered the decisions challenged on appeal.6 (See Haworth v. Superior Court (2010) 50 Cal.4th 372, 379, fn. 2.) II. Principles of Law A. Unfair Competition Law “California’s [UCL] defines ‘unfair competition’ to mean and include ‘any unlawful, unfair or fraudulent business act or practice and unfair, deceptive, untrue or misleading advertising and any act prohibited by [the false advertising law (§ 17500 et seq.)].’ ” (Kasky v. Nike, Inc. (2002) 27 Cal.4th 939, 949.) “The UCL’s purpose is to protect both consumers and competitors by promoting fair competition in commercial markets for goods 6 We also note in their response brief, the People point out that after the trial court reduced defendants’ appeal bond from $33 million to $7 million, defendants voluntarily paid the entire $22.4 million civil penalty award, as evidenced by a copy of a July 2022 acknowledgement of satisfaction of judgment that defendants elected to include in their appellants’ appendix. Because the satisfaction of judgment was not before the trial court when it issued the challenged rulings, it is outside the scope of our review and we do not rely on it. 21 and services.” (Ibid.; see Cel-Tech Communications, Inc. v. Los Angeles Cellular Telephone Co. (1999) 20 Cal.4th 163, 180 (Cel-Tech) [the UCL “has as a major purpose ‘the preservation of fair business competition’ ”].) The scope of the UCL’s substantive provisions is broad. “ ‘By proscribing “any unlawful” business practice, “section 17200 ‘borrows’ violations of other laws and treats them as unlawful practices” that the unfair competition law makes independently actionable.’ ” (Abbott Laboratories v. Superior Court (2020) 9 Cal.5th 642, 651‒652 (Abbott Labs).) Relevant here, “the UCL specifically provides that any practice that violates the FAL is also prohibited by the UCL.” (Nationwide Biweekly Administration, Inc. v. Superior Court (2020) 9 Cal.5th 279, 305 (Nationwide).) “ ‘[T]he section was intentionally framed in its broad, sweeping language, precisely to enable judicial tribunals to deal with the innumerable “ ‘new schemes which the fertility of man’s invention would contrive.’ ” ’ ” (Cel-Tech, supra, 20 Cal.4th at p. 181.) “To that end, the Legislature has created a scheme of overlapping enforcement authority. Section 17204 provides that actions for relief under the UCL may be prosecuted ‘by the Attorney General or a district attorney or by a county counsel authorized by agreement with the district attorney in actions involving violation of a county ordinance, or by a city attorney of a city having a population in excess of 750,000, or by a city attorney in a city and county or, with the consent of the district attorney, by a city prosecutor in a city having a full-time city prosecutor in the name of the people of the State of California upon their own complaint or upon the complaint of a board, officer, person, corporation, or association, or by a person who has suffered injury in fact and has lost money or property as a result of the unfair competition.’ ” (Abbott Labs, supra, 9 Cal.5th at p. 652.) 22 The relief afforded by the UCL is “equitable in nature; damages cannot be recovered.” (Korea Supply Co. v. Lockheed Martin Corp. (2003) 29 Cal.4th 1134, 1144.) Private litigants and public prosecutors are authorized to seek restitution or an injunction. (Cortez v. Purolator Air Filtration Products Co. (2000) 23 Cal.4th 163, 177 [restitution]; People v. Superior Court (Jayhill Corp.) (1973) 9 Cal.3d 283, 286 (Jayhill) [restitution]; In re Tobacco II Cases (2009) 46 Cal.4th 298, 319 [injunction].) However, the UCL “does not mandate restitutionary or injunctive relief [even] when an unfair business practice has been shown.” (Cortez, at p. 180, quoting § 17203.) Public prosecutors, including the Attorney General, “ ‘have an additional tool to enforce the state’s consumer protection laws: civil penalties. “Any person who engages, has engaged, or proposes to engage in unfair competition shall be liable for a civil penalty not to exceed two thousand five hundred dollars ($2,500) for each violation, which shall be assessed and recovered in a civil action brought in the name of the people of the State of California by the Attorney General” ’ or other specified public prosecutors.” (People v. Johnson & Johnson (2022) 77 Cal.App.5th 295, 317 (Johnson & Johnson); § 17206, subd. (a).) Unlike the other forms of relief available under the UCL, “[c]ivil penalties ‘are mandatory once a violation of [the UCL] is established, and a penalty must be imposed for each violation.’ ” (Johnson & Johnson, at p. 317.) B. False Advertising Law “The FAL ‘broadly prohibit[s] false or misleading advertising, declaring that it is unlawful for any person or business to make or distribute any statement to induce the public to enter into a transaction “which is untrue or misleading, and which is known, or which by the exercise of reasonable care should be known, to be untrue or misleading.” ’ ” (Johnson & Johnson, supra, 23 77 Cal.App.5th at p. 317.) Like the UCL, the FAL is “ ‘intended to preserve fair competition and protect consumers from market distortions.’ ” (Kwikset Corp. v. Superior Court (2011) 51 Cal.4th 310, 331.) Also like the UCL, “ ‘the governing substantive standard of the FAL— prohibiting advertising that is “untrue or misleading” [citation]—is set forth in broad and open-ended language that is intended to permit a court of equity to reach any novel or creative scheme of false or misleading advertising that a deceptive business may devise.’ ” (Johnson & Johnson, supra, 77 Cal.App.5th at pp. 317–318, quoting Nationwide, supra, 9 Cal.5th at p. 308.) “ ‘[T]he FAL prohibits “ ‘not only advertising which is false, but also advertising which[,] although true, is either actually misleading or which has a capacity, likelihood or tendency to deceive or confuse the public.’ [Citation.] Thus, to state a claim under either the UCL or the false advertising law, based on false advertising or promotional practices, ‘it is necessary only to show that “members of the public are likely to be deceived.” ’ ” ’ ” (Johnson & Johnson, at p. 318, quoting Nationwide, at p. 309.) Other aspects of the FAL also mirror the UCL. Like the UCL, “FAL actions may be brought by the Attorney General, designated public prosecutors, or ‘any person who has suffered injury in fact and has lost money or property’ as a result of a violation of the FAL.” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 318, citing § 17535.) Further, “[t]he remedies and penalties provided for in the UCL and FAL generally are cumulative to each other and to remedies and penalties available under other laws.” (Johnson & Johnson, at p. 318, citing §§ 17205, 17534.5.) As with the UCL, “the Attorney General and other public prosecutors may seek civil penalties not to exceed $2,500 for each violation of the FAL.” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 318, citing § 17536, 24 subd. (a).) Because the penalties available under each statute are cumulative, “conduct that violates both the UCL and FAL can result in separate penalties of up to $2,500 for each UCL violation and for each FAL violation.” (Johnson & Johnson, at p. 318, citing §§ 17205, 17534.5; see People v. Toomey (1984) 157 Cal.App.3d 1, 22 (Toomey) [by authorizing cumulative penalties, §§ 17205 and 17534.5 evince a legislative intent to “allow . . . double fines”].) And “[t]he duty to impose a penalty for each violation is mandatory.” (People v. First Federal Credit Corp. (2002) 104 Cal.App.4th 721, 728 (First Federal).) The provisions of the UCL and FAL that govern the determination of civil penalties are identical. Both statutes provide: “The court shall impose a civil penalty for each violation of this chapter. In assessing the amount of the civil penalty, the court shall consider any one or more of the relevant circumstances presented by any of the parties to the case, including, but not limited to, the following: the nature and seriousness of the misconduct, the number of violations, the persistence of the misconduct, the length of time over which the misconduct occurred, the willfulness of the defendant’s misconduct, and the defendant’s assets, liabilities, and net worth.” (§§ 17206, subd. (b), 17536, subd. (b).) As this text makes clear, “ ‘[t]he amount of the penalty depends in the first instance on the number of violations committed.’ ” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 352, quoting People ex rel. Kennedy v. Beaumont Investment, Ltd. (2003) 111 Cal.App.4th 102, 127 (Beaumont).) And because “[t]he UCL and FAL do not specify what constitutes a single violation, . . . courts must decide what amounts to a violation on a case-by-case basis.” (Johnson & Johnson, at p. 352, citing Beaumont, at p. 128.) 25 III. The Judgment Must be Reduced to Eliminate Penalties Imposed for Expired FAL Violations Defendants’ first contention is that the judgment must be reduced to correct an error in the trial court’s penalty calculations. They contend the court erroneously included in its violation count FAL violations committed outside the applicable statute of limitations. We agree. As noted above, the parties entered a tolling agreement that became effective February 6, 2013. The FAL has a three-year statute of limitations (Code Civ. Proc., § 338, subd. (h)). Thus, the earliest date defendants could be held liable for FAL violations was February 6, 2010. (See Johnson & Johnson, supra, 77 Cal.App.5th at p. 310, fn. 3 [under tolling agreement with effective date of October 17, 2012, the earliest date the defendant could be held liable for FAL violations was October 17, 2009].) Although we start from the presumption that a trial court decision is correct, this presumption yields if error is “ ‘affirmatively shown.’ ” (Denham v. Superior Court (1970) 2 Cal.3d 557, 564 (Denham).) Here, we are persuaded the trial court inadvertently included stale FAL violations in its violation count. The court found “a total of 1,243,099 misleading calls,” each of which violated the UCL and FAL. On the one hand, early in the statement of decision, the court did acknowledge the limitations period applicable to each statute. On the other hand, the appendix attached to the statement of decision shows the court’s total penalty count included calls reaching back to March 1, 2009. In assessing penalties, the court did not state that it excluded from its calculation FAL violations committed before February 6, 2010. And the total penalty was $22,375,782, which works out to $18 per misleading call, or $9 for each statutory violation. Although the court did not explain 26 how it calculated the total penalty, defendants point out that if one assumes the penalty excluded the expired FAL violations, then the per-violation amount becomes an unlikely sum calculated down to the millionth of one cent. For these reasons, we are persuaded the trial court erred. Defendants have asked us to reduce the judgment by $939,279, but their calculation assumes 11.25 excess months of FAL violations. This overstates the error, because there was not a full quarter month of FAL violations (i.e., seven out of 28 days in February) beyond the 11 months inadvertently included by the court. Defendants bear the burden of demonstrating error, and they have not established the judgment contains a full 11.25 excess months of FAL violations. (See Denham, supra, 2 Cal.3d at p. 564 [where error is not affirmatively shown, judgment will be affirmed].) Defendants have not provided us with a copy of the tolling agreement.7 However, according to the statement of decision, the agreement had an effective date of February 6, 2013. In Johnson & Johnson, we interpreted a tolling agreement with an effective date of October 17, 2012, as making October 17, 2009 the earliest date FAL violations were actionable. (Johnson & Johnson, supra, 77 Cal.App.5th at p. 310, fn. 3.) Applying that reasoning here, the earliest date FAL violations were actionable under the tolling agreement was February 6, 2010. Thus, only the FAL violations from March 1, 2009 through February 5, 2010 (11.18 months rather than 11.25 months) must be eliminated from the judgment. 7 The record on appeal includes a 2017 amendment to the tolling agreement, but not the original agreement. 27 According to the court’s findings, there were 426,734 misleading calls between March 1, 2009 and December 31, 2012, which works out to an average of 9,277 misleading calls per month during this period. This, in turn, leads to the conclusion the FAL violation count includes 103,717 excess calls (11.18 times 9,277) which, when multiplied by $9 per call, equals $933,453. We will reduce the judgment by this amount. IV. The Trial Court Did Not Err in the Methods by Which It Counted UCL and FAL Violations Next, defendants raise two arguments challenging the trial court’s counting of violations of the UCL and FAL: the first based on the court’s reliance on experts who used statistical methods to extrapolate the total number of misleading calls, and the second based on the court’s decision to count each misleading call as a separate violation. “The UCL and FAL do not specify what constitutes a single violation, so courts must decide what amounts to a violation on a case-by-case basis.” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 352.) The trial court’s imposition of civil penalties is reviewed for an abuse of discretion. (Ibid.) “The abuse of discretion standard is not a unified standard; the deference it calls for varies according to the aspect of a trial court’s ruling under review. The trial court’s findings of fact are reviewed for substantial evidence, its conclusions of law are reviewed de novo, and its application of the law to the facts is reversible only if arbitrary and capricious.” (Haraguchi v. Superior Court (2008) 43 Cal.4th 706, 711–712, fn. omitted (Haraguchi).) Here, defendants argue only that the trial court applied the wrong legal standard in assessing penalties. Their position is that the court thereby abused its discretion because “ ‘[t]he scope of discretion always resides in the 28 particular law being applied.’ ” (Horsford v. Board of Trustees of California State University (2005) 132 Cal.App.4th 359, 393 (Horsford).) Thus, we consider whether defendants have met their burden of demonstrating that the trial court committed any of the legal errors they identify. (See Haraguchi, at pp. 711–712 [degree of deference accorded must be guided by the “aspect of a trial court’s ruling under review”].) A. There Was No “Impermissible Trial by Formula” First, defendants argue the trial court erred by tallying violations using an “impermissible trial by formula,” a procedure they imply was held in Duran v. U.S. Bank National Assn. (2014) 59 Cal.4th 1 (Duran) to violate due process. They observe that the trial court found 1,243,099 violations by extrapolating from a “set of just 126 calls.” They suggest this deprived them of an opportunity to litigate their defenses, because “[w]hether any particular statement was misleading is a highly individualized inquiry that should have been evaluated in the context of the specific coursework and degree programs, financial aid and debt issues, and professional aspirations under discussion (not to mention the follow-up calls made and written materials provided to the same prospective students that were never considered).” We are not persuaded. In essence, defendants’ argument is that the trial court erred as a matter of law by relying on statistical evidence to determine the number of violations, because to do so is to conduct an “impermissible trial by formula” prohibited in Duran. (See Horsford, supra, 132 Cal.App.4th at p. 393 [“ ‘[t]he scope of discretion always resides in the particular law being applied’ ”].) Defendants’ argument falls well short of its goal, however, because Duran contains no such holding. The phrase “trial by formula” appears once within the Duran opinion, and only within the title of a law review article cited by 29 the Duran court. (See Duran, supra, 59 Cal.4th at p. 42, citing Lahav, The Case for “Trial by Formula” (2012) 90 Tex. Law Rev. 571, 630.) Thus, rather than accurately reflecting the holding of Duran, the phrase is not even dicta. What Duran did disapprove was the adoption by a trial court of a “seriously flawed” trial plan in a wage and hour class action based on alleged misclassification of employees as exempt from overtime requirements. (Duran, supra, 59 Cal.4th at p. 33.) The trial court in Duran devised a statistical sampling plan, “without following a valid statistical model developed by experts,” for use in the liability phase of the bench trial. (Id. at pp. 12, 33, italics added.) The court “jettison[ed] all statistically grounded criticism of the sampling plan the court itself created” and then “adamantly adhered to [its] methodology, rejecting substantial expert criticism” (id. at p. 49), including the opinions of the plaintiffs’ statistics expert and the defense statistics expert (id. at pp. 22–23). The court also “adamantly refused to admit relevant evidence . . . outside the sample group.” (Id. at p. 33.) Our high court reversed the judgment, explaining that “[a] trial plan that relies on statistical sampling must be developed with expert input and must afford the defendant an opportunity to impeach the model or otherwise show its liability is reduced.” (Id. at p. 13.) Defendants’ assertion that the use of statistical evidence to determine aggregate liability constitutes an “impermissible trial by formula,” or constitutes legal error per se, thus finds no footing in Duran. Instead, the converse is true. The Duran court observed that “representative testimony and sampling may sometimes be appropriate tools.” (Duran, supra, 59 Cal.4th at p. 35.) Other authorities are in accord. “Representative testimony, surveys, and statistical analysis all are available as tools to render manageable determinations of the extent of liability.” (Brinker Restaurant 30 Corp. v. Superior Court (2012) 53 Cal.4th 1004, 1054 (conc. opn. of Werdegar, J.), citing, inter alia, Bell v. Farmers Ins. Exchange (2004) 115 Cal.App.4th 715, 749–755 [upholding as consistent with due process the use of surveys and statistical analysis to measure a defendant’s aggregate liability]; see People v. JTH Tax, Inc. (2013) 212 Cal.App.4th 1219, 1249–1251 [endorsing trial court’s calculation of the number of false print and television advertisements in violation of UCL and FAL based on percentage of gross circulation figures for advertisements and Nielsen ratings for television advertisements].) Defendants neither argue nor establish that errors like those that occurred in Duran were committed by the trial court here. Far from creating its own statistical sampling plan, the trial court in this case relied on the People’s experts Dr. Lucido (a subject matter expert in college admissions) and Dr. Siskin (a statistician). The court found both of these experts’ testimony credible. Over the course of more than three pages of its statement of decision, the court recited the particulars of these experts’ opinions and methods. In considering and ultimately adopting the People’s experts’ methodology, the court expressly stated it was adhering to the following guidance from Duran: “If sampling is used to estimate the extent of a party’s liability, care must be taken to ensure that the methodology produces reliable results. With input from the parties’ experts, the court must determine that a chosen sample size is statistically appropriate and capable of producing valid results within a reasonable margin of error.” (Duran, supra, 59 Cal.4th at p. 42.) As for defendants’ remaining arguments, while they assert the court “extrapolat[ed] from a set of just 126 calls,” they do not develop this assertion into a cognizable argument that the extrapolation resulted in a statistically 31 unreliable result as a matter of law. Similarly, defendants’ contention that “[w]hether any particular statement was misleading . . . should have been evaluated in the context of the specific coursework and degree programs, financial aid and debt issues, and professional aspirations under discussion” is offered as a bare assertion without any citation to supporting legal authority or evidence in the record. Because defendants fail to develop these points further or cite authority in support of them, they are forfeited on appeal. (Allen v. City of Sacramento (2015) 234 Cal.App.4th 41, 52 (Allen) [“When legal argument with citation to authority is not furnished on a particular point, we may treat the point as forfeited and pass it without consideration. . . . We are not required to examine undeveloped claims or to supply arguments for the litigants.”]; Paterno v. State of California (1999) 74 Cal.App.4th 68, 106 [“An appellate court is not required to examine undeveloped claims, nor to make arguments for parties.”].) Further, defendants’ claim about what “should” have been done at trial begs the question: who should have done it? Under Duran, defendants had the procedural right to present to the trial court any of the factual contentions they now raise. But they do not argue they were deprived of this opportunity. Instead, they seek to have us reverse the trial court’s decision based on facts of which we are not independently aware (the importance of considering “specific coursework and degree programs,” for example) and for which they have provided no support. For the reasons we have already identified, we decline to do so. In their reply brief, defendants argue for the first time on appeal that “Johnson & Johnson underscores the . . . error” committed by the trial court here. We disagree. In the passage of Johnson & Johnson on which they rely, this court upheld defendant Ethicon’s substantial evidence challenge to the 32 trial court’s finding that oral marketing communications by Ethicon’s sales representatives were likely to deceive doctors. (Johnson & Johnson, supra, 77 Cal.App.5th at pp. 340–342.) We reached that conclusion based on our inability to locate record support showing the content of any of Ethicon’s oral marketing communications. (See id. at pp. 329 [applying substantial evidence standard of review], 341 [explaining the lack of evidentiary support for trial court’s finding].) Here, defendants did not raise in their opening brief a substantial evidence challenge to the trial court’s factual findings (nor have they done so in their reply brief). Any such challenge has therefore been forfeited. (Tisher v. California Horse Racing Board (1991) 231 Cal.App.3d 349, 361 (Tisher) [holding that the “failure to raise an issue in the[ ] opening brief waives the issue on appeal].) Further, our holding in Johnson & Johnson was expressly limited to the determination “there was insufficient evidence in this case regarding the substance of Ethicon’s oral marketing communications.” (Johnson & Johnson, at p. 342.) Johnson & Johnson thus does not stand for the proposition that the number of oral communications violating the UCL and FAL can never be aggregated by using statistical methods developed by experts. B. The Trial Court Did Not Abuse Its Discretion by Counting Each Misleading Call as a Separate Statutory Violation Next, defendants contend the trial court erred by counting each deceptive phone call placed by an admissions counselor as a separate violation of the UCL and FAL. Relying primarily on Jayhill, supra, 9 Cal.3d 283, Toomey, supra, 157 Cal.App.3d 1, and People v. Superior Court (Olson) (1979) 96 Cal.App.3d 181 (Olson), defendants contend other courts have generally calculated violations “on a ‘per victim’ basis rather than a ‘per [mis]representation’ one.” Pointing to trial testimony of two former Ashford employees, defendants assert that admissions counselors spoke to each victim 33 on average “a ‘half-dozen’ times.” Defendants argue we should therefore “divide the number of violations found below by six” and reduce the judgment accordingly. The People respond that California courts have “rejected the ‘contention that the number of violations must always be calculated on a “per victim” rather than a “per act” basis.’ ” (See Beaumont, supra, 111 Cal.App.4th at p. 129.) They argue the trial court’s approach in this case was valid and consistent with a per-advertisement methodology used in Morse, supra, 21 Cal.App.4th at pages 272 to 274 and Johnson & Johnson, supra, 77 Cal.App.5th at page 355. They also contend the trial record does not support the six-calls-per-victim average defendants rely on in seeking a reduced judgment. We agree with the People the trial court did not abuse its discretion by counting each misleading call as a separate violation. Defendants’ argument questions the trial court’s application of law to the facts. Such a decision “is reversible only if arbitrary and capricious.” (Haraguchi, supra, 43 Cal.4th at pp. 711–712.) First, defendants fail to show that the trial court acted outside the scope of its discretion by counting each misleading call as a separate violation of the UCL and FAL. No statutory authority prohibited it from doing so, because “[t]he UCL and FAL do not specify what constitutes a single violation.” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 352.) In arguing that violations should have been counted on a per-victim basis, defendants rely in part on Jayhill, supra, 9 Cal.3d 283, a case involving misrepresentations by door-to-door encyclopedia sellers. There, our high court said, “we believe the Legislature intended that the number of violations is to be determined by the number of persons to whom the 34 misrepresentations were made, and not by the number of separately identifiable misrepresentations involved.” (Id. at p. 289.) But in subsequent decisions courts have limited Jayhill to its facts. “ ‘[I]t is unlikely that Jayhill intended to establish a test for determining the number of violations applicable to all situations.’ ” (Beaumont, supra, 111 Cal.App.4th at p. 129 [citing cases].) It is now accepted that “determining what qualifies as a single violation ‘depends on the type of violation involved, the number of victims and the repetition of the conduct constituting the violation—in brief, the circumstances of the case.’ ” (Ibid.; see Nationwide, supra, 9 Cal.5th at p. 314 [stating that in FAL cases trial courts have “equitable authority . . . in determining . . . the number of violations for which a defendant may properly be held responsible”]; Johnson & Johnson, supra, 77 Cal.App.5th at p. 352 [“courts must decide what amounts to a violation on a case-by-case basis”].) Thus, no legal authority required the trial court to count violations on a per-victim rather than a per-communication basis. Second, defendants fail to persuade us the trial court’s decision to count each misleading call as a violation was arbitrary under the facts of this case. Each phone call in the violation count was found to contain at least one deceptive statement. The court did not act outside the scope of its authority by considering each deceptive sales call to be a separate statutory violation, because this approach has already been endorsed. (See e.g., Jayhill, supra, 9 Cal.3d at p. 289.) Defendants’ challenge to the violation count assumes there were deceptive calls included in the violation count that were made to the same student. As the People observe, the violations were calculated using a statistical sample, and defendants fail to establish the sample included repeat encounters. 35 Third, even if we were to assume the violation count does include repetitive deceptive calls to the same student, we would find no abuse of discretion. In Morse, supra, 21 Cal.App.4th at pages 272 to 274, the Court of Appeal rejected a contention that it was an abuse of discretion to base an award of FAL penalties “ ‘on the number of solicitations mailed, rather than the number of people who received and read the solicitations, or who responded to the solicitations.’ ” The defendant was an attorney who mailed false and misleading solicitations to homeowners offering assistance in filing homestead declarations. The appellate court explained the mailers were “highly individualized solicitations” that included information specific to the recipient that the defendant obtained from public records. (Id. at pp. 273‒ 274.) The court concluded that “[u]nder these circumstances, it is reasonable to assume that the Legislature contemplated a penalty for each direct mailing.” (Id. at p. 274.) In Johnson & Johnson, the trial court counted each IFU (instruction for use) and marketing communication that contained a false or misleading statement as a separate violation of the UCL and FAL. (Johnson & Johnson, supra, 77 Cal.App.5th at p. 353.) On appeal, defendant Ethicon argued the trial court abused its discretion by “adopting a per-communication methodology to calculate the total number of violations.” (Ibid.) Rejecting this argument, we explained the deceptive statements “were substantively targeted to well-defined groups of people.” (Id. at p. 355.) The IFUs were specifically directed to doctors considering whether to implant Ethicon’s device or preparing to do so, and the marketing communications “were explicitly written to appeal to those same doctors[.]” (Ibid.) Further, the IFUs and marketing communications “were sent, displayed, or made available only to those same limited audiences, not the broader general 36 public.” (Ibid.) “Given the highly targeted nature of Ethicon’s communications,” we concluded the trial court “reasonably found each IFU and marketing communication represented a gain or opportunity to gain for Ethicon.” (Ibid.) Morse and Johnson & Johnson support the conclusion the trial court was within its rights to consider each deceptive phone call a separate violation of the UCL and FAL, even if we assume the violation count includes repeat encounters. Each phone call was an oral marketing communication comparable to the individual mailers in Morse and the IFUs and marketing communications in Johnson & Johnson. The court’s findings support the inference the communications addressed the personal financial, career, and educational concerns of individuals who were contemplating enrolling in Ashford. Moreover, the court’s findings demonstrate the admissions counselors who delivered the communications were under pressure to increase enrollment numbers. The “highly individualized” (Morse, supra, 21 Cal.App.4th at p. 273) and “highly targeted” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 355) nature of the communications, together with the pressure placed on admissions counselors, support counting each deceptive call as a separate violation. Although defendants argue that it was not clear in Morse or Johnson & Johnson whether the same communications reached the same person, the reasoning of Johnson & Johnson nevertheless supports counting repeat communications as separate violations, on the theory each successive deceptive encounter “represent[s] a gain or opportunity to gain for [defendants].” (Johnson & Johnson, at p. 355.) Nor do the other cases cited by defendants persuade us the trial court abused its discretion by counting each misleading call as a separate violation. Although in Toomey, supra, 157 Cal.App.3d at page 23, the appellate court 37 endorsed the trial court’s use of a “ ‘per victim’ ” basis to calculate violations, finding it “manifestly reasonable in the case of telephone solicitations,” the court did not discuss or reject other approaches. (Agnew v. State Bd. of Equalization (1999) 21 Cal.4th 310, 332 [“ ‘It is axiomatic . . . that a decision does not stand for a proposition not considered by the court.’ ”].) And while defendants cite Olson, supra, 96 Cal.App.3d at page 196 as an example of a case in which the Attorney General argued for a per-victim approach, the prior argument of a party in another case has no bearing on our decision in this case. V. The Trial Court Did Not Violate Principles of Extraterritoriality Defendants next challenge the civil penalty on the ground it violates principles of “extraterritoriality” to the extent it encompasses statements that caused harm outside California. Based on the factual finding that defendants’ misconduct “emanated from California” for the vast majority of the statutory period, the trial court included in its violation count all deceptive calls placed by defendants nationwide, both to residents of California and to residents of other states. The court stated that by doing so it was following “the well-established rule that the UCL extends to conduct that emanates from California even if victims reside out of state.” Explaining its determination that defendants’ conduct “emanated from California” and therefore fell within this rule, the court cited evidence that defendants were headquartered in San Diego, and Ashford’s former presidents, admissions counselors, and the majority of admissions employees worked in San Diego. Although defendants moved their headquarters to Arizona in 2019, many of their top executives remained 38 in San Diego, and defendants continued to employ admissions counselors and compliance staff in San Diego. Defendants do not dispute the trial court’s finding that their “misconduct emanated from California.” Instead, they assert their “decision to communicate alleged misrepresentations from California does not justify applying the UCL to nonresident call recipients when their harm (if any) occurred only out of state.” They argue that including such calls to nonresidents in the violation count violated principles of extraterritoriality. Questions of law and the application of law to undisputed facts are subject to de novo review. (People v. Superior Court (Sahlolbei) (2017) 3 Cal.5th 230, 234; Boling v. Public Employment Relations Bd. (2018) 5 Cal.5th 898, 912.) Although our standard of review is de novo, defendants bear the burden of demonstrating error. (Doe v. Lawndale Elementary School Dist. (2021) 72 Cal.App.5th 113, 139, fn. 8 (Lawndale Elementary); Denny v. Arntz (2020) 55 Cal.App.5th 914, 920 (Denny).) As we explain, they fail to meet that burden here. Section 17500, which defines the conduct prohibited by the FAL, makes it illegal to make or disseminate deceptive statements “from this state.” (See § 17500 [providing it is unlawful for a business or any of its employees “to make or disseminate or cause to be made or disseminated from this state . . . any statement . . . which is untrue or misleading, and which is known, or which by the exercise of reasonable care should be known, to be untrue or misleading” (italics added)].) Under section 17200, unfair competition within the meaning of the UCL means and includes “any act prohibited by” section 17500. (See § 17200 [unfair competition shall mean and include “any act prohibited by Chapter 1 (commencing with Section 17500) of Part 3 of Division 7 of the Business and Professions Code”].) 39 Wershba v. Apple Computer, Inc. (2001) 91 Cal.App.4th 224 (Wershba) (cited by the trial court in the statement of decision) was a class action under the UCL against a computer company (Apple) for discontinuing free technical support it had promised customers in brochures advertising and accompanying its products. (Id. at pp. 230–231, disapproved on other grounds in Hernandez v. Restoration Hardware (2018) 4 Cal.5th 260, 269.) The Wershba court ruled that the UCL could constitutionally be applied to the claims of nonresident class members because Apple was a California corporation with its principal place of business in California, substantial numbers of class members were located in California, and the brochures at issue were prepared in and distributed from California. (Id. at pp. 241–242.) In doing so, it relied in part on section 17500, which it said “expressly applies to claims by out-of-state class members deceived by representations ‘disseminated from’ the state of California.” (Wershba, at p. 242.) The Wershba court also relied on Clothesrigger, Inc. v. GTE Corp. (1987) 191 Cal.App.3d 605 (Clothesrigger), another case the trial court cited in support of its finding the UCL extends to defendants’ conduct that emanates from California even if victims reside out of state. In Clothesrigger, the court found sufficient contacts with California such that California law could constitutionally be applied to fraud and unfair business practices claims of nonresident class members where the defendant did business in California, its principal offices were in California, a significant number of class members were California residents, and the employees and agents who prepared the misleading literature at issue did so in California. (Clothesrigger, at p. 613; Wershba, supra, 91 Cal.App.4th at pp. 241–242; see Wershba, at p. 242 [“ ‘Thus the alleged fraudulent misrepresentations 40 forming the basis of the claim of every [class member] nationwide emanated from California.’ ” (Italics added.)].) Defendants do not cite section 17500 or challenge the scope of conduct it proscribes as exceeding the Legislature’s territorial jurisdiction. They also do not argue that the trial court acted outside the scope of its statutory authority under sections 17200 and 17500 by ruling that misleading statements emanating from California violated the UCL and FAL. Nor do they contend that Wershba or Clothesrigger were wrongly decided. And, as we have mentioned, they also do not challenge the trial court’s factual finding that their deceptive conduct “emanated from California.” Any such claims have therefore been forfeited. (Tisher, supra, 231 Cal.App.3d at p. 361 [holding that the “failure to raise an issue in the[ ] opening brief waives the issue on appeal].) Instead, defendants offer a series of assertions intended to demonstrate that the trial court violated principles of extraterritoriality by penalizing them for their deceptive calls to non-California residents. Their first argument is based on “due process concerns.” They cite State Farm Mutual Automobile Insurance Co. v. Campbell (2003) 538 U.S. 408 (State Farm) for the proposition that a state does not “have a legitimate concern in imposing punitive damages to punish a defendant for unlawful acts committed outside of the State’s jurisdiction.” (See State Farm, at p. 421.) Next, they cite Philip Morris USA v. Williams (2007) 549 U.S. 346 for the proposition that “the Constitution’s Due Process clause forbids a State to use a punitive damages award to punish a defendant for injury that it inflicts upon nonparties or those whom they directly represent.” (See Philip Morris, at p. 353.) Finally, they contend that “[t]hose same due process concerns are implicated here, where the Attorney General represents the citizens of this State, not others.” 41 They cite sections 17206, subdivision (a), and 17536, subdivision (a), in support of this last contention. Defendants’ first argument is underdeveloped and ultimately unpersuasive. Although we do not disagree with the proposition for which defendants quote State Farm, they fail to show it has any application here. Defendants have not challenged the court’s factual finding that their misleading statements emanated from California, a finding that tracks the language of section 17500. (See § 17500 [providing it is unlawful for a business or any of its employees “to make or disseminate or cause to be made or disseminated from this state . . . any statement . . . which is untrue or misleading, and which is known, or which by the exercise of reasonable care should be known, to be untrue or misleading” (italics added)].) Accordingly, they have not established that their misconduct was “committed outside [California’s] jurisdiction.” (See State Farm, supra, 538 U.S. at p. 421.) As for defendants’ reliance on Philip Morris, that case’s prohibition against using punitive damages to punish a defendant for injuries it inflicts on nonparties does not apply in this context. This is an action to enforce the UCL and FAL brought by a public prosecutor, not a private lawsuit brought by an injured plaintiff seeking recovery of damages. Harm is not an element of an action for civil penalties under the UCL and FAL. (See §§ 17206, subd. (a), 17536, subd. (a); Johnson & Johnson, supra, 77 Cal.App.5th at pp. 317–318; People v. Dollar Rent-A-Car Sys. (1989) 211 Cal.App.3d 119, 131.) A court may impose civil penalties under these statutes “without individualized proof of reliance, deception and injury if it is convinced that such a remedy is necessary to deter [the defendant’s] unfair practice.” (Dollar Rent-A-Car, at p. 131, italics added.) The penalty in this case was 42 imposed to punish defendants for violating the UCL and FAL, not for injuring nonresidents (or nonparties). And defendants’ assertion about the Attorney General’s representation of citizens of this state appears to reflect a misunderstanding of the Attorney General’s role in an action for civil penalties under the UCL and FAL. In support of their assertion, defendants cite sections 17206, subdivision (a), and 17536, subdivision (a), which provide that civil penalties may be recovered in civil actions brought by the Attorney General “in the name of the people of the State of California.” (Italics added.) Although defendants do not explain their reasoning, they seem to interpret the italicized text to mean the Attorney General acts as a representative of California victims in such actions. This is incorrect. When the Attorney General initiates an action asserting violations of the UCL and FAL, he does so under his constitutional authority as the chief law enforcement officer of the state. (See Cal. Const., art. V, § 13; People v. Eubanks (1996) 14 Cal.4th 580, 589 [“The public prosecutor ‘ “is the representative not of any ordinary party to a controversy, but of a sovereignty.” ’ ”]; Abbott Labs, supra, 9 Cal.5th at p. 659 [“The UCL does not undermine the Attorney General’s constitutional role as California’s chief law enforcement officer.”].) “ ‘[I]n the absence of any legislative restriction, [the Attorney General] has the power to file any civil action or proceeding directly involving the rights and interests of the state, or which he deems necessary for the enforcement of the laws of the state, the preservation of order, and the protection of public rights and interest.’ ” (D’Amico v. Board of Medical Examiners (1974) 11 Cal.3d 1, 14–15.) To the extent defendants perceive a legislative restriction on the Attorney General’s authority in the statutory text stating that civil penalties shall be recovered in an action “brought in the name of the people of the State 43 of California” (§§ 17206, subd. (a), 17536, subd. (a)), this, too, is incorrect. Under Government Code section 100, “[t]he sovereignty of the state resides in the people thereof, . . . all writs and processes shall issue in their name.” (Gov. Code, § 100, subd. (a).) This statute further specifies that “[t]he style of all process shall be ‘The People of the State of California,’ and all prosecutions shall be conducted in their name and by their authority.” (Gov. Code, § 100, subd. (b).) In enacting Business and Professions Code sections 17206 and 17536, the Legislature was presumably aware of these procedural conventions. (People v. Overstreet (1986) 42 Cal.3d 891, 897 [“the Legislature is deemed to be aware of existing laws and judicial decisions in effect at the time legislation is enacted”].) Thus, by specifying in sections 17206, subdivision (a), and 17536, subdivision (a), that actions for civil penalties shall be brought “in the name of the people of the State of California,” the Legislature did not signal an intent to limit the Attorney General’s authority to prosecute violations of the UCL or FAL to cases involving false or misleading statements directed at California residents. Next, defendants rely on People ex rel. Lockyer v. R.J. Reynolds Tobacco Co. (2004) 116 Cal.App.4th 1253, in which we reversed a sanction award that was based on a company’s national advertising budget. However, the sanction in that case was based on breach of a consent decree with a limited geographic scope. (See id. at pp. 1261, 1283–1284 [defendant violated consent decree and settlement agreement by targeting tobacco advertising “to youth within California”].) This case does not help defendants. In defendants’ remaining arguments, they appear in part to conflate the location where prohibited conduct occurred with the location where the resulting harm occurred. Within the same paragraph of their opening brief, for example, they first frame the issue raised by their appeal as “whether the 44 State [can] collect penalties for conduct occurring outside its borders”; two sentences later, they characterize the same issue as “whether the State can collect penalties for harm that occurred outside its borders.” (Original italics omitted; our italics added.) But defendants do not squarely contend, much less present a developed argument, that their misconduct occurred outside our state’s borders. Their failure to present a developed argument on this point, as opposed to a general assertion, forfeits any such claim of error. (See People v. Stanley (1995) 10 Cal.4th 764, 793 [deeming waived a defendant’s argument on appeal that merely “makes a general assertion, unsupported by specific argument”]; People v. Hardy (1992) 2 Cal.4th 86, 150 [rejecting due process claim because other than a brief mention of the argument, appellant failed to expand on the issue].) And defendants’ assertion that the presumption against extraterritoriality prohibits recovery of penalties for harm occurring outside the state is simply incorrect. “The presumption against extraterritoriality is one against an intent to encompass conduct occurring in a foreign jurisdiction in the prohibitions and remedies of a domestic statute.” (Diamond Multimedia Systems, Inc. v. Superior Court (1999) 19 Cal.4th 1036, 1059, fn. 20.) In Diamond Multimedia, the defendants allegedly made fraudulent statements and omissions in California that induced nonresidents to make out-of-state purchases of securities. The issue was whether, when conduct violating Corporations Code section 25400 is committed in California, nonresidents who purchase their stock out of state may recover damages under Corporations Code section 25500. (Diamond Multimedia, at pp. 1040– 1041.) Rejecting an argument that permitting nonresidents to recover damages for such purchases would run afoul of the presumption against extraterritoriality, our high court explained because Corporations Code 45 sections 25400 and 25500 “provide[d] a remedy for a wrongful act committed in California,” there was no potential “for a ‘clash’ between California law and that of any other state or of the United States.” (Diamond Multimedia, at p. 1059, fn. 20.) The same reasoning applies here and leads to the conclusion that the presumption of extraterritoriality is not implicated by permitting the recovery of civil penalties for harm occurring outside California where, as here, the conduct violating the relevant law was committed in California. Next, defendants appear to fault the trial court for relying on Wershba and Clothesrigger because they are “about choice-of-law principles.” Their argument overlooks that Wershba and Clothesrigger were based on choice-of- law and constitutional principles. (See Wershba, supra, 91 Cal.App.4th at pp. 241–244; Clothesrigger, supra, 191 Cal.App.3d at pp. 612–616.) But even if we were to assume that defendants are correct and Wershba and Clothesrigger are only choice-of-law cases, this would not matter since a court cannot hold the UCL instead of another state’s law governs the claims of out- of-state residents aggrieved by a California defendant’s deceptive statements unless the UCL does, in fact, apply to such claims. Finally, defendants contend that Sullivan v. Oracle Corp. (2011) 51 Cal.4th 1191 (Sullivan) “reinforces that the Attorney General cannot recover penalties for students located outside of California.” Sullivan does not assist defendants. In Sullivan, the California Supreme Court held the UCL did not apply to a California employer’s asserted failure to comply with the federal Fair Labor Standards Act (FLSA) (29 U.S.C. § 201 et seq.) by paying nonresident employees for overtime worked in other states. (Sullivan, at pp. 1206–1209.) The problem in Sullivan was that there was no evidence the employer’s conduct that violated federal overtime laws was committed in 46 California. The only conduct committed inside California was the employer’s misclassification of the nonresident workers, but adopting an erroneous classification policy was not unlawful “in the abstract.” (Sullivan, at p. 1208.) What was unlawful was the failure to pay overtime when due, but the record did not establish where payment occurred. (Ibid.) The Court held that under these circumstances, the UCL did not apply. Sullivan’s holding is not apposite here because the underlying violations in this case arose under the FAL, not the FLSA, and this case does not involve a lack of evidence defendants committed the misconduct violative of the FAL inside California. In their reply brief, defendants contend their development of policies for admissions counselors and “decision to communicate alleged misrepresentations from California” were “abstract,” just like the employer’s classification decision in Sullivan. They claim the harm to out-of-state residents in this case “was akin to the out-of-state failure to pay in Sullivan.” But their attempt to analogize this case to Sullivan falls short, because it ignores that the trial court in this case found the actual law violations occurred in this state, a factual finding defendants have not challenged. (See Sullivan, supra, 51 Cal.4th at p. 1208, fn. 10 [distinguishing Wershba and Clothesrigger as cases in which “the unlawful conduct that formed the basis of the out-of-state plaintiffs’ claims . . . , and that justified the application of California law to resolve those claims, occurred in California”].) For all these reasons, defendants fail to persuade us that the trial court violated principles of extraterritoriality by awarding penalties based on their deceptive statements to non-residents of California. 47 VI. The Penalty Is Not Excessive Defendants’ last challenge is that the penalty violated the prohibitions against excessive fines and grossly disproportionate punishments embraced by the excessive fines and due process clauses of the United States Constitution, as well as the corresponding provisions of the California Constitution. (See U.S. Const., 8th & 14th Amends.; Cal. Const., art. I, §§ 7, 17.) The California Supreme Court has held that appellate courts need not separately analyze whether a penalty is constitutionally excessive and whether it violates due process. (People ex rel. Lockyer v. R.J. Reynolds Tobacco Co. (2005) 37 Cal.4th 707, 728 (Reynolds).) In Reynolds, the court examined the constitutionality of an award of civil penalties using a multi- factor test borrowed from United States v. Bajakajian (1998) 524 U.S. 321 (Bajakajian), a seminal Eighth Amendment case. Under this test, whether a penalty is unconstitutionally excessive depends on “four considerations: (1) the defendant’s culpability; (2) the relationship between the harm and the penalty; (3) the penalties imposed in similar statutes; and (4) the defendant’s ability to pay.” (Reynolds, at p. 728, citing Bajakajian, at pp. 337‒338.) Here, defendants directly challenge only the second and fourth Reynolds factors. They argue, first, that the penalty should be reduced because it is not reasonably related to the harm proven at trial, and second, that it should be reduced because it is excessive given Zovio’s financial condition. We analyze defendants’ arguments with respect to these factors and then summarize our conclusions with respect to the remaining factors. 48 A. Standard of Review In reviewing the excessiveness of a penalty, “ ‘we accept the trial court’s factual findings unless clearly erroneous and determine de novo whether the fine is excessive.’ ” (Johnson & Johnson, supra, 77 Cal.App.5th at p. 359; see City and County of San Francisco v. Sainez (2000) 77 Cal.App.4th 1302, 1313 (Sainez) [“our review of the ruling on the constitutional question is independent judgment, or de novo [citation], but with deference to underlying factual findings, which we review for substantial evidence, viewing the record in the light most favorable to the ruling”].) Because the factual findings in the statement of decision are unchallenged on appeal, we presume they are correct. (Adhav, supra, 37 Cal.App.5th at p. 969; PR/JSM, supra, 180 Cal.App.4th at p. 1486 [trial court’s unchallenged factual findings presumed correct on appeal].) Defendants contend certain factual findings were omitted from the statement of decision. Specifically, they contend the trial court failed to explain how the penalty was based on the victims’ harm, and that it made no findings with respect to their ability to pay. In the ordinary course, trial courts are not required to provide written findings of fact after a bench trial. (Code Civ. Proc., § 632.) Where factual findings are not expressly made, under the doctrine of implied findings an appellate court will fill in the gap by presuming the missing finding was decided in favor of the prevailing party. (In re Marriage of Arceneaux (1990) 51 Cal.3d 1130, 1133.) The implied finding is then reviewed under the substantial evidence standard. (Fladeboe v. American Isuzu Motors Inc. (2007) 150 Cal.App.4th 42, 60 (Fladeboe).) Sections 632 and 634 of the Code of Civil Procedure set forth actions a party can take to avoid the doctrine of implied findings. These statutes 49 “describe a two-step process for avoiding implied factual findings. First, a party must request a statement of decision pursuant to Code of Civil Procedure section 632. [Citation.] Second, if the trial court issues a statement of decision, a party claiming omissions or ambiguities in the factual findings must bring the omissions or ambiguities to the trial court’s attention.” (Fladeboe, supra, 150 Cal.App.4th at p. 59.) An appellant who fails to follow these rules and thereafter complains on appeal that there are factual findings missing from the statement of decision will find itself subject to the doctrine of implied findings. (Id. at pp. 59–60.) We first consider defendants’ claim that the trial court failed to explain how the penalty was based on the evidence of the victims’ harm. According to defendants, not only were such findings omitted from the statement of decision, but we also cannot infer the assertedly missing findings. They contend they succeeded in avoiding the doctrine of implied findings because they complied with Code of Civil Procedure section 634 by filing a motion for new trial in which they objected that the degree of harm shown to have been experienced by the victims was not in proportion with the penalty, rendering the penalty unconstitutionally excessive. We disagree. Defendants did not file a request for statement of decision as required by Code of Civil Procedure section 632. And in their new trial motion, defendants argued, in part, that the penalty should be reduced because it was not sufficiently related to the trial evidence of harm. They did not suggest they desired any additional factual findings on harm or object that the statement of decision omitted findings on this issue. This is not compliance with Code of Civil Procedure section 634. (Tyler v. Norton (1973) 34 Cal.App.3d 717, 725–726.) Accordingly, “[b]ecause defendants do not show that express findings on [this] factor[ was] legally required or requested for inclusion in the statement of 50 decision, we are free to follow our usual appellate course and imply . . . findings to support the ruling.” (Sainez, supra, 77 Cal.App.4th at p. 1313.) As mentioned, defendants also assert that the trial court made “no findings” regarding their ability to pay the penalty. Once again, they contend the assertedly missing findings8 cannot be inferred in favor of the judgment because they complied with Code of Civil Procedure section 634 by raising the omission in their new trial motion. Again, we disagree. Defendants did not request findings on Zovio’s ability to pay the penalty in their new trial motion. Rather, they argued that the trial court failed to make findings on Zovio’s net worth and that this omission required it to vacate the judgment and reconsider its decision. But even if we were to treat this as sufficient to satisfy Code of Civil Procedure section 634 with respect to the full gamut of financial facts relevant to defendants’ ability to pay, avoiding the doctrine of implied findings is a two-step process. Defendants ignore the first step— filing a request that identifies the controverted issues to be included in the statement of decision (see Code Civ. Proc., § 632)—and they do not argue they 8 We also disagree that the statement of decision is devoid of findings with respect to Zovio’s financial condition or reflects a failure to consider Zovio’s finances in setting the amount of civil penalties. A statement of decision must disclose the “factual and legal basis for [the trial court’s] decision as to each of the principal controverted issues at trial[.]” (Code Civ. Proc., § 632.) It need not discuss every subsidiary issue. (Wolfe v. Lipsy (1985) 163 Cal.App.3d 633, 643, disapproved on other grounds in Droeger v. Friedman, Sloan & Ross (1991) 54 Cal.3d 26, 36–39.) The failure to make an express finding on a material issue is harmless when the missing finding may reasonably be found to be implicit in other findings. (See, e.g., Rojas v. Mitchell (1996) 50 Cal.App.4th 1445, 1450 (Rojas); Wolfe, at pp. 643–644.) As we discuss below, the court made findings with respect to Zovio’s financial state. Implicit in these findings is a determination Zovio had the ability to pay the penalty. 51 should be excused from complying with this requirement. We are not obligated to develop defendants’ arguments for them.9 (See Benach v. County of Los Angeles (2007) 149 Cal.App.4th 836, 852 [“It is not our place to construct theories or arguments to undermine the judgment and defeat the presumption of correctness.”].) In any event, the impact of the doctrine of implied findings with respect to the ability to pay factor is mitigated “by the relative absence at trial of disputed issues of fact.” (LSREF2 Clover Property 4, LLC v. Festival Retail Fund 1, LP (2016) 3 Cal.App.5th 1067, 1076 (LSREF2).) As we shall discuss, the trial evidence relevant to Zovio’s financial state consisted of the testimony of a single witness as well as several SEC filings. The individual data points established by this evidence are undisputed. And “[w]e review legal conclusions from an established set of facts independently.” (Ibid.) B. Consideration of the Reynolds Factors Does Not Demonstrate the Penalty Is Excessive 1. Contrary to Defendants’ Arguments, the Penalty Is Not Insufficiently Related to Victims’ Harm Defendants argue the penalty does not bear a reasonable relationship to the trial evidence of the harm caused by their deceptions. Referring to the People’s request for $222,119 in restitution on behalf of the nine testifying 9 Defendants do not contend the trial court committed reversible procedural error by purportedly failing to make express findings on harm or ability to pay. (See F.P. v. Monier (2017) 3 Cal.5th 1099, 1114, 1116 [erroneous failure to comply with statutory requirements pertaining to issuance of a statement of decision subject to harmless error review].) Accordingly, any such claim has been forfeited. (Scheenstra v. California Dairies, Inc. (2013) 213 Cal.App.4th 370, 403 [“fundamental rule of appellate review is that the appellant must affirmatively show prejudicial error”]; Tisher, supra, 231 Cal.App.3d at p. 361 [the “failure to raise an issue in the[ ] opening brief waives the issue on appeal”].) 52 student victims, they assert that no harm was established at trial other than these victims’ economic losses. They further contend the penalty should not be permitted to exceed the four-to-one ratio between compensatory and punitive damages typically authorized in punitive damages cases. Therefore, since the People only proved defendants’ misconduct caused $222,119 in “actual harm,” the penalty must be reduced to $888,476—a “harm” to penalty ratio of one to four. We are not persuaded. Defendants’ claim that the penalty must not exceed a single-digit ratio of the monetary loss proven at trial relies on BMW of North America, Inc. v. Gore (1996) 517 U.S. 559 (Gore) and progeny (e.g., State Farm, supra, 538 U.S. at p. 425). Gore established three guideposts for courts to consider when determining whether an award of punitive damages violates the federal due process clause: (1) the degree of reprehensibility of the defendant’s misconduct; (2) the disparity between the actual or potential harm suffered by the plaintiff and the punitive damages award [expressed as a ratio of compensatory damages to punitive damages]; and (3) the difference between the punitive damages awarded and the civil penalties authorized or imposed in comparable cases. (Gore, at pp. 575–585.) For a number of reasons, we disagree that the Gore test applies in this context. First, in Reynolds, our Supreme Court expressly adopted the Bajakajian test to analyze the constitutionality of a civil penalty. (Reynolds, supra, 37 Cal.4th at p. 728.) “On federal questions, intermediate appellate courts in California must follow the decisions of the California Supreme Court, unless the United States Supreme Court has decided the same question differently.” (Correia v. NB Baker Electric, Inc. (2019) 32 Cal.App.5th 602, 619, citing Auto Equity Sales, Inc. v. Superior Court (1962) 57 Cal.2d 450, 455.) And the United States Supreme Court “has not 53 suggested that the Gore guideposts should extend to constitutional review of statutory damage awards.” (Sony BMG Music Entertainment v. Tenenbaum (1st Cir. 2013) 719 F.3d 67, 70 (Sony BMG).) Second, as the People observe, other appellate courts have found it inappropriate to use tests formulated for punitive damage awards to assess the proportionality of a civil penalty. (See People ex rel. Bill Lockyer v. Fremont Life Ins. Co. (2002) 104 Cal.App.4th 508, 520 (Fremont) [declining to apply Gore guideposts to assess the constitutionality of a civil penalty because the Gore guideposts “address the propriety of a punitive damage award, not a civil penalty”]; First Federal, supra, 104 Cal.App.4th at pp. 731– 732 [declining to apply to a civil penalty award the holding of Adams v. Murakami (1991) 54 Cal.3d 105, 108–109 (Murakami) requiring evidence of a defendant’s financial condition to be presented by a plaintiff in a punitive damages case].) As the First Federal court observed, there are fundamental differences between statutory penalties and punitive damages that make it illogical to analyze an award of statutory penalties using tests formulated to evaluate the disproportionality of a punitive damage award. (First Federal, at p. 732.) For example, “civil penalties under the [UCL and FAL] are only imposed in actions brought by public law enforcement officials, not by private litigants.” (Ibid.) And the concern about juror passion or prejudice affecting a punitive damage award “is absent in UCL cases because there is no right to a jury trial in such cases. [Citation.] Runaway jury verdicts cannot occur when there is no jury to inflame.” (Id. at pp. 732–733.) Also, unlike a punitive damages case, in an action for civil penalties under the UCL and FAL, the People do not have to prove victims suffered actual damages. (Toomey, supra, 157 Cal.App.3d at p. 23.) Only the violation of the statute is necessary to justify civil penalties. (Ibid.) And both 54 the UCL and FAL provide, “The court shall impose a civil penalty for each violation of this chapter.” (§§ 17206, subd. (b), 17536, subd. (b), italics added.) If a violation is proven, it is error for the court not to impose penalties. (People v. Custom Craft Carpets (1984) 159 Cal.App.3d 676, 686.) Tethering the constitutionality of such civil penalties to the amount of actual damages proven at trial makes little sense given that the court is required to assess penalties even in a case with no established damages. (See Sony BMG, supra, 719 F.3d at pp. 70–71 [explaining that applying the second Gore guidepost “cannot logically apply to an award of statutory damages under the Copyright Act[ because it] requires a comparison between the award and the harm to the plaintiff, but a plaintiff seeking statutory damages under the Copyright Act need not prove actual damages”].) These differences make the Gore guideposts a poor fit for assessing the disproportionality of a civil penalty. (See Sony BMG, at p. 70 [declining to apply the Gore guideposts to an award of statutory damages under the Copyright Act; observing that “[t]he concerns regarding fair notice to the parties of the range of possible punitive damage awards, which underpin Gore, are simply not present in a statutory damages case where the statute itself provides [fair] notice of the scope of the potential award”].) Moreover, courts analyzing the constitutionality of a civil penalty have not required a quantified sum of victim damages akin to a jury’s compensatory damages award in order to uphold the penalty as not excessive. In Johnson & Johnson, supra, 77 Cal.App.5th 295, for example, in determining that defendant Ethicon’s misleading statements and omissions in medical device instructions caused “severe harm,” we reasoned in part that “Ethicon harmed all consumers by depriving their doctors of material information necessary to counsel patients and forcing patients to make 55 potentially life-altering decisions about their health and well-being based on this same false or incomplete information.” (Id. at pp. 360, 361 [analyzing excessiveness of a $302 million civil penalty according to the Reynolds factors].) We further reasoned that “an especially unlucky subset of patients experienced more severe harm.” (Id. at p. 360.) We did not scrutinize the record for evidence of the monetary value of the harm caused by Ethicon’s unlawful conduct. And in People v. Braum (2020) 49 Cal.App.5th 342, $6 million in civil penalties was imposed against a landlord who violated a city’s zoning code by leasing properties for use as medical marijuana dispensaries. In deciding the penalties were not excessive, the court reasoned the landlord “increased the risk of the harm the City was endeavoring to enjoin and abate” by allowing unpermitted uses to continue on the leased properties, and that the landlord’s failure to comply with the city’s requests “dictated that the amount of penalties necessary to achieve the City’s legitimate enforcement goals would be substantial.” (Id. at p. 362.) The court did not identify any particular expenses or losses caused by the landlord’s conduct or limit itself to considering only such harm. For all of these reasons, we decline to apply the second Gore guidepost here, or to restrict our consideration of the harm caused by defendants’ conduct to losses that were quantified at trial. Setting aside defendants’ misplaced reliance on Gore, we do not otherwise find the harm disproportionate to the penalty. In considering this question, we presume the facts set forth in the statement of decision are correct. (See PR/JSM, supra, 180 Cal.App.4th at p. 1486 [trial court’s unchallenged factual findings presumed correct on appeal].) Further, as discussed above, because defendants have not demonstrated full compliance 56 with Code of Civil Procedure sections 632 and 634, “we are free to follow our usual appellate course and imply [any omitted factual] findings to support the ruling.” (Sainez, supra, 77 Cal.App.4th at p. 1313.) We start by rejecting defendants’ position that the only harm established at trial was the $222,119 in economic losses suffered by their victims. Consistent with the reasoning of cases like Johnson & Johnson and Braum, in addition to the monetary losses the nine testifying victims were proven to have suffered, we also consider the noneconomic harm caused by defendants’ deceptions. The statement of decision contains factual findings relevant to nonmonetary harm. The trial court expressly credited Dr. Lucido’s expert testimony and gave it “significant weight.” Lucido testified the choice of whether and where to go to college is a “momentous decision” that “sets [forth] the trajectory for [students’] lives and their family” and often represents “the biggest cost of anything [students] purchase, other than perhaps a home.” When admissions counselors do not give students information “accurately and faithfully,” students “can make a choice that is not in their best educational interest, and that ultimately might result in a cost that they didn’t expect or inability to achieve their goal.” The court also found defendants enrolled students with “ ‘complex’ and ‘difficult lives,’ which ‘heightens’ the need for accurate college advising.” It further found that “[a] typical Ashford bachelor’s degree . . . cost[s] between $40,000 and $60,000 during the statutory period” yet “[o]nly a quarter of Ashford students graduate [citations], and many default on their student loans[.]” The court credited the testimony of the nine student victims, each of whom described how their admission counselors’ misrepresentations led them to make decisions that frustrated their goals, deprived them of years of lost time, cost 57 them financially, and had other detrimental consequences. And the deceptive statements to which these victims succumbed were not isolated instances: the court found “significant similarities between the deception identified by Dr. Lucido in the phone calls and the stories of the testifying victims.” In their reply brief, defendants assert that any nonmonetary harm suffered by victims cannot be considered because there was no expert evidence “regarding the monetary value that ought to be assigned to the evidence of nonmonetary harm.” But they offer only this bare assertion, and they make it without any supporting legal authority. Accordingly, the argument is forfeited. (Allen, supra, 234 Cal.App.4th at p. 52 [“When legal argument with citation to authority is not furnished on a particular point, we may treat the point as forfeited and pass it without consideration. . . . We are not required to examine undeveloped claims or to supply arguments for the litigants.”].) Defendants also try to minimize the harm caused by admissions counselors’ misrepresentations by presenting a one-sided version of the evidence that understates the effects of their illegal conduct. They assert, for example, that Alison T.—the aspiring teacher who spent over $50,000 on her Ashford education degree after she was falsely told it would carry over to Pennsylvania—was able to more than double her prior salary after graduating by working as a phlebotomist.10 Yet they ignore Alison’s testimony that she was “[o]verwhelmed” by her student loan balance of 10 Alison initially worked for a private school after graduating from Ashford, but soon lost that job when the school closed and could not find another one without a teaching license, forcing her to pursue a different career. Although she testified her salary at this school was $22,000, double her prior salary of $11,000, defendants fail to cite evidence she was employed there long enough to make up for the cost of her Ashford degree. 58 $59,000 and found it “devastating” to have gone so far into debt “for a degree [she’s] not using.” Moreover, Alison testified she did not need a bachelor’s degree to work as a phlebotomist, which means the salary increase she obtained by changing career paths cannot be credited to her Ashford degree. Defendants also state that another former student, Rene W., suffered no cognizable losses because while her classes cost $4,500, “her loans were forgiven before she made any payments to Ashford.” We have reviewed Rene’s testimony and find defendants’ summary to be highly misleading. Rene (a single mother whose children “were on free and reduced lunch”) testified she enrolled in Ashford after her admissions counselor said her prior college credits would transfer and she would have no out-of-pocket costs because everything would be covered by financial aid. She trusted the information her admissions counselor provided and enrolled after he “calmed all [her] fears and made [her] feel safe.” After completing three classes, she learned her admissions counselor had enrolled her in the wrong degree program. She realized she had not received financial aid after Ashford sent her a bill showing she owed over $4,500 for the classes she had completed. Nothing in the excerpts of testimony cited by defendants shows Rene received loans or that any such loans were forgiven. To the contrary, Rene testified she had not yet paid Ashford “any of the balance that they say [she] owe[s]” because she “believe[s] that financial aid was granted” and “feel[s] like [she was] completely taken advantage of.” (Italics added.) In their reply brief, defendants transform the meaning of this testimony by selectively quoting Rene as saying she “did not ‘pa[y] Ashford any of the balance’ because ‘financial aid was granted.’ ” (Italics added.) Similarly, defendants highlight evidence Ashford forgave another victim, Loren E., a $3,900 payment she owed. But they ignore that the trial 59 court found Loren “reached her lifetime loan limit” before graduating, leaving her with “massive debt but no degree,” and that her student ledger showed she took out more than $20,000 in student loans. Defendants’ reliance on evidence that favors them, rather than the judgment, is contrary to the principles that guide our review (Sainez, supra, 77 Cal.App.4th at p. 1313; Adhav, supra, 37 Cal.App.5th at p. 969), and their record mischaracterizations undermine their credibility and detract significantly from the persuasiveness of their arguments. The parties dispute whether the harm suffered by the nine students who testified at trial can be extrapolated to the full population of Ashford victims. We need not resolve this issue. Even if we assume the trial court could not properly draw such an inference, the court still could reasonably conclude the harmful consequences these students experienced showed the seriousness of admissions counselors’ misrepresentations. Further, the court could reasonably find defendants harmed all students by providing them deceptive and misleading information about the financial consequences of pursuing a degree at Ashford, ability to transfer credits into and out of Ashford, pace of Ashford coursework, and career prospects upon graduating, all of which was information material to “potentially life-altering decisions.” (See Johnson & Johnson, supra, 77 Cal.App.5th at p. 360.) That the misconduct was directed at a vulnerable population with a heightened need for accurate advising supports the conclusion this harm was substantial. We find the penalty imposed to be reasonably related to the substantial harm caused by defendants’ deceptive conduct. The total assessment works out to just $9 per violation, a conservative, even lenient, punishment considering the nature of the deceptions perpetrated and risks created by each misleading call. Toomey, supra, 157 Cal.App.3d 1, a case involving 60 undisclosed restrictions on the use of coupons sold as part of gambling packages, is one of the few cases in which such a minimal penalty was imposed. (See id. at pp. 7–8.) There, the court imposed a penalty of $1 per violation for each of 150,000 violations of the UCL and FAL, for a total fine of $150,000. (Id. at p. 23.) And yet the deceptions in this case were material to decisions far more significant and potentially life-altering than whether to buy gambling coupons. We conclude that defendants’ misleading and untrue statements were substantially harmful and that the penalty imposed is reasonably related to this harm. This factor weighs strongly against a finding of excessiveness. 2. Defendants Fail to Demonstrate Zovio’s Inability to Pay the Penalty Defendants also contend the penalty is excessive given Zovio’s financial condition.11 They claim the trial court made “no findings” relative to Zovio’s ability to pay. They assert that Zovio had “nearly $70 million in liabilities, . . . just ‘$31.6 million in unrestricted cash,’ ” and an “operating . . . ‘loss of around $18 million’ ” in September 2021, and that Zovio’s net worth in December 2020 was only $6.05 million.12 They cite a series of punitive damages cases for the propositions that “severe penalties” are in “contravention of due process of law” (St. Louis, I.M. & S.R. Co. v. Williams 11 In advancing their ability-to-pay challenge, defendants make factual assertions based on an exhibit filed in support of their motion for relief from the bond requirement. As we have discussed, we disregard facts derived from materials that were not before the trial court when it issued its statement of decision or considered the merits of defendants’ new trial motion. (See Glassman, supra, 90 Cal.App.5th at p. 1307 [matters not before trial court when it ruled on the orders challenged on appeal may not be considered].) 12 The trial took place in November and December of 2021. 61 (1919) 251 U.S. 63, 64–65) and the goal of punitive damages “is to deter, not to destroy” (Murakami, supra, 54 Cal.3d at p. 112). But they primarily rely on an older punitive damages case, Storage Services v. Oosterbaan (1989) 214 Cal.App.3d 498, 515 (Storage Services), in which the court observed that “punitive damages awards are generally not allowed to exceed 10 percent of the defendant’s net worth.” On the strength of Storage Services, defendants claim the penalty must be reduced to $605,000, because this figure represents 10 percent of Zovio’s asserted net worth of $6.05 million. Defendants’ arguments include the passing assertion that the trial court made “no findings” relative to Zovio’s ability to pay. As noted, we disagree with this characterization of the court’s statement of decision. (See fn. 7, ante.) To explain why, we provide some additional background. The trial evidence of Zovio’s financial state consisted of the testimony of one defense witness, Jim Smith, Jr., Ashford’s former senior vice president of finance and the current CFO of UAGC, as well as several SEC filings. Smith testified that as of the quarter ending September 2021, Zovio had approximately $31.6 million of unrestricted cash and equivalents, current liabilities of approximately $70 million, year-to-date revenues of $200 million, and had reported an operating loss of around $18 million for the nine months ending September 30, 2021. He also admitted on cross-examination that in September 2020, before Zovio sold Ashford to UAGC, Zovio had cash and cash equivalents of $86 million, and that as of the time of trial Zovio was currently providing enrollment services to UAGC and was receiving payments from UAGC. Details of Zovio’s 2020 sale of Ashford to UAGC were set forth in an SEC filing (an August 2020 8-K form) admitted as an exhibit at trial. This exhibit showed that Zovio had entered into a purchase agreement with 62 UAGC pursuant to which Zovio had agreed to pay a total of $54 million in cash to UAGC, and to transfer Ashford to UAGC, in exchange for consideration of $1. In return, Zovio would provide enrollment services to UAGC, and UAGC would pay Zovio fees equal to its direct costs ($10 million to $25 million annually), as well as a fluctuating percentage of 15.5 to 19.5 percent of UAGC’s tuition and fees revenue, with the percentage subject to increase or decrease if UAGC’s tuition and fee revenue fell below or exceeded $440 million. In closing arguments to the trial court, the People asked the court to impose a $75 million penalty. Defense counsel argued a penalty of this size “would realistically ruin this company.” He did not attempt to quantify Zovio’s net worth, but he argued that defendants’ “assets, liabilities, and net worth . . . , as you heard from Jim Smith” were such that Zovio could not pay a $75 million penalty. In its statement of decision, the trial court did not explicitly address defendants’ contention that Smith’s testimony supported the conclusion Zovio would be “ruin[ed]” by a multimillion dollar civil penalty. However, the findings the court did make can reasonably be interpreted as a rejection of this claim, at least as it pertained to a penalty in the reduced amount ultimately imposed. (See e.g., Rojas, supra, 50 Cal.App.4th at p. 1450 [assertedly omitted findings “may reasonably be found to be implicit in other findings”].) In determining the amount of the penalty, the court first listed all of the statutory penalty factors, including “the defendant’s assets, liabilities, and net worth” (§§ 17206, subd. (b), 17536, subd. (b)), and stated it “considered the factors” as well as “the entire record” (italics added) and decided that $22,375,782 in statutory penalties is “reasonable and supported by the evidence.” 63 Although the trial court did not detail all of the relevant evidence in this segment of the statement of decision, it did so elsewhere. Early in its decision, the court found “Ashford has generated hundreds of millions of dollars for Zovio annually” (a fact it derived from Zovio’s form 10-K SEC filings from 2010 through 2020). It also observed that “[i]n December 2020 . . . . [i]n exchange for paying $54 million to ‘sell’ Ashford to UAGC, Zovio will now receive 15.5-19.5% of UAGC’s tuition revenue for the next 7-15 years. (Ex. 735.0002-3.)” The only relevance of these facts and evidence was to the financial factors in the civil penalties provisions of the UCL and FAL. (See §§ 17206, subd. (b), 17536, subd. (b) [“the defendant’s assets, liabilities, and net worth”].) Moreover, these provisions do not require the court to consider every financial factor, but only those the court finds relevant. (See Fremont, supra, 104 Cal.App.4th at p. 528 [trial courts assessing the amount of civil penalties under §§ 17206, subd. (b), and 17536, subd. (b), must consider “among other factors and if relevant, ‘the defendant’s assets, liabilities, and net worth’ ”].) Inferably, the court included this financial information in its written decision because it found the information relevant and supportive of the penalty it imposed. For these reasons, we disagree that the statement of decision lacks findings relevant to Zovio’s ability to pay the penalty.13 Further, at the hearing on defendants’ new trial motion, the trial court confirmed it had “look[ed] at the whole record” and had relied on the terms of the 2020 transaction, including that Zovio had agreed to pay UAGC over $50 million to sell Ashford in return for a future income stream, in setting the amount of 13 As discussed above (see fn. 8, ante), defendants do not argue the trial court committed procedural error by supposedly omitting findings on Zovio’s ability to pay the penalty. 64 the penalty. By denying the motion, the court implicitly found the penalty did not exceed Zovio’s ability to pay. On appeal, defendants fail to establish that the trial court reached the wrong conclusion. As we have discussed, because the facts are undisputed, we determine de novo whether they demonstrate Zovio’s inability to pay. (See LSREF2, supra, 3 Cal.App.5th at p. 1076 [“We review legal conclusions arising from an established set of facts independently.”].) Although our review is independent, defendants bear the burden of demonstrating error. (Lawndale Elementary, supra, 72 Cal.App.5th at p. 139, fn. 8; Denny, supra, 55 Cal.App.5th at p. 920.) Once again, we are not persuaded by their arguments. First, defendants rely heavily, if not exclusively, on the 10-percent-of- net-worth benchmark identified in Storage Services, supra, 214 Cal.App.3d at page 515. But Storage Services is an older punitive damages case, and after it was decided, our high court expressly declined to adopt net worth as the standard for determining a defendant’s ability to pay. (See Murakami, supra, 54 Cal.3d at p. 116, fn. 7 [“Defendant in this case contends the best measure of his ability to pay is his net worth. . . . We decline at present, however, to prescribe any rigid standard for measuring a defendant’s ability to pay.”].) Further, courts reviewing the constitutionality of punitive damage awards have agreed “ ‘[n]et worth’ is subject to easy manipulation and . . . should not be the only permissible standard.” (Lara v. Cadag (1993) 13 Cal.App.4th 1061, 1065, fn. 3; accord Zaxis Wireless Communications v. Motor Sound Corp. (2001) 89 Cal.App.4th 577, 580–582 [affirming a punitive damages award of $300,000, although the defendant had a negative net worth of $6.3 million]; Bankhead v. ArvinMeritor, Inc. (2012) 205 Cal.App.4th 68, 79–84 [affirming punitive damages award of $4.5 million against a claim it was 65 grossly disproportionate to defendant’s negative net worth of $1.023 billion].) The viability of the 10-percent-of-net-worth benchmark adopted in punitive damages cases like Storage Services has been questioned, in part due to concern that net worth can be an untrustworthy standard. (See Bankhead, at pp. 80‒84.) And in the civil penalties context, courts have similarly declined to look at only select metrics when determining the constitutionality of a penalty. In Sainez, supra, 77 Cal.App.4th at page 1320, the court rejected an invitation to adopt such a narrow approach. There, landlords who were penalized for code violations in one of their several buildings argued the court should consider only their equity or rental proceeds from the affected building when deciding whether the penalty was excessive. (Id. at pp. 1319–1320.) The court declined the invitation, explaining that “[l]imiting the penalty to a certain proportional share of the equity (market value minus the purchase price or remaining debt) in one property could mask the fact that the equity had served as security or a down payment for other properties.” (Ibid.) It would also create “[a] second and related problem,” namely “inviting defendants to use manipulative financing.” (Id. at p. 1320.) These problems could “artificially insulate a defendant from penalties and remove all effective deterrence.” (Ibid.) The Sainez court instead concluded that “consideration of a defendant’s overall financial status is proper.” (Ibid., citing People ex rel. Van de Kamp v. Cappuccio, Inc. (1988) 204 Cal.App.3d 750, 765 [proper to consider defendant’s “financial condition,” including salary and annual gross income, in due process review of penalty].) For these reasons, we are not persuaded to use a 10-percent-of-net- worth benchmark, or to rely exclusively on Zovio’s net worth, in determining 66 whether the penalty is excessive. Instead, we will consider Zovio’s overall financial status. Here, the trial court relied in part on the terms of Zovio’s 2020 sale of Ashford in determining the penalty was not excessive. We agree it is reasonable to consider this transaction. If a defendant could obtain reduced penalties by strategically structuring the sale of its affected business so as to transfer existing assets from the present to the future, this “could artificially insulate [the] defendant from penalties.” (Sainez, supra, 77 Cal.App.4th at p. 1320.) According to Smith, in September 2020, before Ashford was sold, Zovio had $86 million in unrestricted cash and cash equivalents. The penalty imposed by the court ($22,375,782) represents 26 percent of Zovio’s unrestricted cash in September 2020. Also, the trial court found Ashford earned Zovio “hundreds of millions of dollars . . . annually.” The penalty constitutes roughly 11 percent of $200 million, the lowest number that can properly be described as “hundreds of millions.” In Sainez, the court compared the landlords’ penalty with two years of rent because they accrued almost two years’ worth of code violations. (Sainez, supra, 77 Cal.App.4th at p. 1317.) If we take that approach here and compare the penalty with 11 years of Zovio’s income from Ashford to reflect that defendants were found to have violated the UCL and FAL over the course of 11 years (from March 2009 to April 2020), the resulting comparison is just one percent. As we have mentioned, on appeal defendants have taken the position their net worth in December 2020 was $6.05 million. They have calculated this net worth figure in a footnote of their opening brief on appeal based on information in a consolidated balance sheet in Zovio’s SEC form 10-K for the fiscal year ending December 31, 2020. The balance sheet reflects that as of 67 December 2020, Zovio had “[t]otal current assets” of $76,833,000 and “[t]otal current liabilities” of $70,783,000. Zovio informs us that the difference in these amounts is $6.05 million, which they then offer as their net worth.14 But the same balance sheet also shows that at that same point in time, Zovio had “[t]otal assets” of $161,306,000 and “[t]otal liabilities” of $102,089,000. Defendants fail to acknowledge these figures, which calculate to a total net worth of $59.22 million, and which better represent Zovio’s overall financial state since they take into account all of its assets and liabilities. The penalty is 37.8 percent of this more comprehensive net worth figure. The rest of the financial metrics relied on by defendants relate to Zovio’s financial state in September 2021, shortly before the trial. Citing Smith’s testimony, they emphasize that Zovio had “nearly $70 million in liabilities, . . . just ‘$31.6 million of unrestricted cash,’ ” and an “operating . . . ‘loss of around $18 million.’ ” But they ignore that Smith also testified Zovio had year-to-date revenues of $200 million and total assets of over $157 million at that same point in time, and that it had begun receiving payments from UAGC. 14 We solicited supplemental briefing from the parties as to whether defendants preserved for appeal their position that Zovio’s net worth in December 2020 was $6.05 million, and that this figure rendered the penalty award unconstitutional. After considering the parties’ briefs, we conclude defendants failed to assert or develop this issue in the trial court. Defendants identify no instance in which they directed the court’s attention to this information or argued it was relevant to the amount of the penalty, whether during the trial or in their motion for new trial. Accordingly, this position has been forfeited. (See Araiza v. Younkin (2010) 188 Cal.App.4th 1120, 1127 [“A party who fails to alert the trial court to an issue that has been left unresolved forfeits the right to raise that issue on appeal.”].) In any event, as we discuss, it also lacks merit. 68 And while the penalty constitutes substantially more than half of Zovio’s $31.6 million in unrestricted cash in September 2021, here we have to consider that the reason Zovio’s cash resources were limited at that time was because it had recently transferred both Ashford and $54 million in cash to UAGC in exchange for $1 and a stream of future income. In other words, it had recently, and voluntarily, depleted its assets. According to the terms of the transaction set forth in the SEC filing relied on by the court, the agreed income was anticipated to be substantially remunerative, $10 million to $25 million in fees representing Zovio’s direct costs plus between 15.5 and 19.5 percent of UAGC’s revenue, depending on whether UAGC’s revenue exceeded or fell below $440 million. Thus, while Zovio had just divested itself of significant assets, it also stood to receive significant future income from its deal with UAGC. If we were to consider only Zovio’s current assets and liabilities but not its anticipated income stream, this would “artificially insulate [defendants] from penalties.” (Sainez, supra, 77 Cal.App.4th at p. 1320.) And while Smith stated Zovio had recently reported $200 million in third-quarter revenues and a $18 million third-quarter operating loss, his summary ended there; he was not asked about Zovio’s projected 2021 performance, and defendants did not introduce evidence indicating Zovio anticipated financial difficulties or would be unable to withstand a judgment far less than the $75 million requested by the People. We believe that when considered as a whole the foregoing figures adequately demonstrate Zovio’s ability to pay the penalty. In Sainez, the court upheld against a due process challenge a penalty that represented 28.4 percent of the defendants’ net worth and 120 percent of rents due for the two- year period of defendants’ civil violations. (See Sainez, supra, 77 Cal.App.4th at p. 1319.) In ruling the penalty also was not an excessive fine, the court 69 noted that “[a] net worth of about $500,000 has been held enough ability to pay a penalty of $353,000.” (Id. at p. 1322, citing U.S. v. Lippert (8th Cir. 1998) 148 F.3d 974, 976, 978.) As a percentage of Zovio’s net worth, the penalty is somewhat greater than the net worth percentage in Sainez, but it is substantially less than the net worth ratio in Lippert (which equates to 70.6 percent). And whereas in Sainez the penalty represented 120 percent of the landlords’ biennial rents, here the penalty represents a far smaller share of Zovio’s gross income, particularly if one considers the 11 years in which it received hundreds of millions of dollars annually through its operation of Ashford while violating the UCL and FAL. Although it had sold Ashford by the time of trial, Zovio was continuing to earn income at this same pace (i.e., over $200 million annually), and the evidence showed it had bargained to receive tens of millions in revenue per year, years into the future, through its transaction with UAGC. Defendants point to Johnson & Johnson, in which this court affirmed a civil penalty of $302 million, which represented less than one half of one percent of Johnson & Johnson’s net worth. (Johnson & Johnson, supra, 77 Cal.App.5th at p. 361.) However, our conclusion in that case that this net worth ratio showed the company had “ample ability to pay” the penalty (ibid.) does not compel the conclusion a higher net worth ratio is excessive as to Zovio. For these reasons, defendants fail to persuade us the penalty exceeds Zovio’s ability to pay, or that they are entitled to a reduced penalty of $605,000. To the contrary, after considering Zovio’s financial status overall, we conclude that the trial evidence sufficiently established its ability to pay the $22,375,782 penalty imposed by the court. The fact that we will be 70 reducing the penalty by almost $1 million further supports this conclusion. Accordingly, this factor also weighs against a finding of excessiveness. C. Consideration of All Four Reynolds Factors Leads Us to Conclude the Penalty Not Excessive So far, we have determined that both the second and fourth Reynolds factors weigh against a finding of excessiveness. (See Reynolds, supra, 37 Cal.4th at p. 728 [the excessiveness factors are “(1) the defendant’s culpability; (2) the relationship between the harm and the penalty; (3) the penalties imposed in similar statutes; and (4) the defendant’s ability to pay”].) This leaves the first and third factors, defendants’ culpability and penalties imposed in similar statutes. As for culpability, defendants assert that the trial court found their overall level of culpability to be “modest.” We disagree with their assessment. The court never used this adjective, and any conclusion the court viewed defendants as having only modest culpability is belied by its express findings. In explaining why it opted to impose the penalty it did, the court stated that defendants “made false and misleading statements to students over a substantial period of time,” a circumstance it plainly considered aggravating. Separately, it also found defendants “knew of misleading statements,” “made scant ‘effort to discourage’ ” them, and “did not refuse to accept the benefits of enrollment based on misrepresentations.” Although the trial court also found there were mitigating factors— defendants’ creation of a compliance program, the work of a third-party monitor who oversaw a 2014 compliance program between defendants and the state of Iowa, and a consent order entered by defendants in 2016— inferably it did not find them so weighty as to warrant a punishment lesser than the one it imposed. 71 The trial court’s other express findings shed light on its reasoning. The three listed mitigating circumstances corresponded to affirmative defenses which defendants raised and the court rejected. Explaining why the existence of defendants’ compliance program did not establish their good faith defense, the court found “it is clear that [d]efendants did not take serious action to prevent or remedy the extensive deception their compliance program identified” and that defendants “[d]id [n]ot [d]emonstrate [g]ood [f]aith” (boldface omitted). The court also found, in connection with defendants’ defense based on opinions of a monitor who oversaw an Iowa compliance program, that the monitorship “lasted only from May 2014 to May 2017 . . . whereas this case ranges from 2009 to 2021”; the monitor’s finding of “no ‘pattern or practice’ of misrepresentations” was contradicted by the People’s evidence; and the monitor “did not investigate the same range of misrepresentations that the People have proven in this case.” As for the 2016 consent order, which was entered as part of defendants’ settlement with the federal Consumer Financial Protection Bureau, the court deemed it irrelevant to liability because it “involved alleged violations of federal law relating to private loans that Zovio made to students which are not at issue in this case,” although it did find that a budgeting tool implemented pursuant to the consent order walked students through “at least some” financial information. Inferably, the court found the actions defendants took (or were forced to take) to address conduct underlying a subset of their offenses insufficient to address the full scope of defendants’ wrongs. Thus, the trial court’s findings do not compel the conclusion defendants’ overall culpability was only modest. Theirs were not mere technical violations of law. (Cf. Bajakajian, supra, 524 U.S. at pp. 337‒338 [defendant’s crime, failure to report transportation of currency, carried a 72 minimal level of culpability where it was “solely a reporting offense”].) Instead, the nature of defendants’ misrepresentations, the overwhelming number of violations, and the length of time over which they were committed, all indicate a serious level of culpability that was not substantially diminished by the mitigators the court identified. We conclude that this factor weighs against a finding of excessiveness. As for the final Reynolds factor, the penalties imposed in similar statutes, the parties identify no other statutes that sanction comparable misconduct. However, the penalty imposed works out to $9 per violation, orders of magnitude below the statutory maximum of $2,500 prescribed by the UCL and FAL (§§ 17206, subd. (a), 17536, subd. (a)), and less than one third of $75 million in penalties the Attorney General requested, all of which weighs heavily against a finding of excessiveness. Although all of the Reynolds factors need not point in the same direction (Johnson & Johnson, supra, 77 Cal.App.5th at p. 361), in this case they do. We therefore conclude the penalty is not excessive or grossly disproportionate. DISPOSITION The judgment is modified as follows: the civil penalties awarded to the People are reduced from $22,375,782 to $21,442,329. The judgment as 73 modified is affirmed. The parties shall bear their own costs on appeal. (Cal. Rules of Court, rule 8.278(a)(5).) DO, J. WE CONCUR: DATO, Acting P. J. KELETY, J. 74