Concurring. — I agree that the trial court committed error in the course of ruling on the class certification motion of named plaintiffs Maria Ayala, Josefina Briseño, Rosa Duran, and Osman Núñez, that remand for further consideration of the motion is necessary, and that affirmance of the Court of Appeal’s judgment is appropriate. The record indicates that the trial court did not adequately consider the extent to which there will be common proof regarding a central factor in determining whether carriers who deliver newspapers for defendant Antelope Valley Newspapers, Inc. (Antelope Valley), are employees or independent contractors: the extent to which Antelope Valley has the right to control the manner and means by which the carriers accomplish their work. The record also suggests that the trial court did not adequately perform the weighing of common and individualized proof necessary to determine whether common issues predominate.
However, in several respects, I question the majority’s legal analysis. I also do not endorse its dicta regarding some of the secondary factors that are relevant to determining whether someone who provides service to another is an employee or an independent contractor. (Maj. opn., ante, at pp. 538-540.) I therefore concur only in the judgment.
I. Factual and Procedural Background.
In December 2008, plaintiffs sued on behalf of a putative class of newspaper carriers, alleging that Antelope Valley improperly treated them as independent contractors instead of employees and improperly denied them various statutory wage and hour protections. The complaint alleged numerous *542violations of our labor laws — including unpaid overtime, unlawful deductions, failure to provide breaks, and failure to reimburse for business expenses — and unfair competition based on those violations. Plaintiffs moved for class certification, contending that the central question in establishing liability— whether carriers are employees or independent contractors — would be resolved through common proof, principally the contracts between Antelope Valley and its carriers.
Antelope Valley opposed certification, arguing in relevant part that there was insufficient commonality regarding proof of its right to control the means and manner by which its carriers accomplish their work, its actual exercise of control, and various secondary factors that, under S. G. Borello & Sons, Inc. v. Department of Industrial Relations (1989) 48 Cal.3d 341, 350 [256 Cal.Rptr. 543, 769 P.2d 399] (Borello), are relevant to determining whether a service provider is an employee or an independent contractor. Antelope Valley further argued that even were the carriers employees, some of the causes of action presented additional unmanageable individual issues that should nevertheless preclude certification.
The trial court denied the certification motion, finding that plaintiffs had failed to show that (1) “common questions of law or fact predominate,” (2) “a class action would be ‘superior’ to individual lawsuits,” or (3) despite the “highly individualized” nature of “the issues affecting the class,” “manageability is achievable through the use of’ various procedural tools, including questionnaires, surveys, and representative sampling. As to the claims still at issue in this appeal, the Court of Appeal reversed, believing that the trial court had based its ruling on “variations in how the carriers performed their jobs,” and finding that “those variations do not present individual issues that preclude class certification.” We then granted Antelope Valley’s petition for review.
II. The Legal Inquiry.
As relevant to this appeal, plaintiffs, as the proponents of certification, had the burden in the trial court to demonstrate that “ ‘questions of law or fact common to the class predominate over the questions affecting the individual members.’ ” (Lockheed Martin Corp. v. Superior Court (2003) 29 Cal.4th 1096, 1104 [131 Cal.Rptr.2d 1, 63 P.3d 913].) “To assess predominance, a court ‘must examine the issues framed by the pleadings and the law applicable to the causes of action alleged.’ [Citation.] It must determine whether the elements necessary to establish liability are susceptible of common proof or, if not, whether there are ways to manage effectively proof of any elements that may require individualized evidence. [Citation.]” (Brinker Restaurant Corp. v. Superior Court (2012) 53 Cal.4th 1004, 1024 [139 *543Cal.Rptr.3d 315, 273 P.3d 513] (Brinker).) Thus, in assessing predominance, courts “must carefully evaluate the nature of the proof’ the parties will present. (Keating v. Superior Court (1982) 31 Cal.3d 584, 622 [183 Cal.Rptr. 360, 645 P.2d 1192].) The “ultimate question” is whether “the issues [that] may be jointly tried, when compared with those requiring separate adjudication, are so numerous or substantial that the maintenance of a class action would be advantageous to the judicial process and to the litigants.” (Collins v. Rocha (1972) 7 Cal.3d 232, 238 [102 Cal.Rptr. 1, 497 P.2d 225].)
The decision to grant or deny a certification motion “rests squarely within the discretion of the trial court” because the trial court is “ ‘ideally situated to evaluate the efficiencies and practicalities of permitting group action.’ ” (Fireside Bank v. Superior Court (2007) 40 Cal.4th 1069, 1089 [56 Cal.Rptr.3d 861, 155 P.3d 268].) Accordingly, reviewing courts “afford” trial court decisions “great deference on appeal, reversing only for a manifest abuse of discretion.” (Ibid.) Under its “narrowly circumscribed” inquiry, a reviewing court generally may not disturb an order denying certification unless “ ‘it is unsupported by substantial evidence’ ” or “ ‘rests on improper criteria . . . or . . . erroneous legal assumptions.’ ” (Brinker, supra, 53 Cal.4th at p. 1022.) In applying this test, a reviewing court “must ‘[p]resum[e] in favor of the [trial court’s] order ... the existence of every fact the trial court could reasonably deduce from the record . . . .’ [Citation.]” (Ibid.)
As we have recognized, the predominance inquiry “may be enmeshed with” issues “affecting the merits of a case.” (Linder v. Thrifty Oil Co. (2000) 23 Cal.4th at pp. 429, 443 [97 Cal.Rptr.2d 179, 2 P.3d 27].) “When evidence or legal issues germane to the certification question bear as well on aspects of the merits, a court may properly evaluate them.” (Brinker, supra, 53 Cal.4th at pp. 1023-1024.) “[I]if the parties’ evidence is conflicting on the issue of whether common or individual questions predominate . . . , the trial court is permitted to credit one party’s evidence over the other’s in determining whether the requirements for class certification have been met — and doing so is not ... an improper evaluation of the merits of the case. [Citations.]” (Dailey v. Sears, Roebuck & Co. (2013) 214 Cal.App.4th 974, 991 [154 Cal.Rptr.3d 480], citing Sav-On Drug Stores, Inc. v. Superior Court (2004) 34 Cal.4th 319, 331 [17 Cal.Rptr.3d 906, 96 P.3d 194] (Sav-On).) To the extent the trial court’s order turns on inferences to be drawn from the facts or on an evaluation of the credibility of conflicting evidence, a reviewing court may not “substitute” its “judgment for the trial court’s.” (Sav-On, supra, at p. 331.) Even at the certification stage, “ ‘questions as to the weight and sufficiency of the evidence, the construction to be put upon it, the inferences to be drawn therefrom, the credibility of witnesses . ., and the determination of [any] conflicts and inconsistency in their testimony are matters for the trial court to resolve.’ [Citation.]” (Id. at p. 334.)
*544Under the complaint, as a prerequisite to recovery, plaintiffs must establish that they are employees of Antelope Valley rather than independent contractors. In litigating the certification motion below, both plaintiffs and Antelope Valley maintained that this issue is governed by the principles and considerations Borello set forth. Consistent with the parties’ arguments, both the trial court and the Court of Appeal applied Borello’s principles in determining whether certification was appropriate. In accordance with Antelope Valley’s petition for review, the issue on which we granted was whether the trial court abused its discretion in finding that, under Borello’s test for determining whether someone is an employee or an independent contractor, common questions of law or fact will not predominate over individual questions.
The issue in Borello was whether, for purpose of workers’ compensation coverage, certain agricultural laborers were employees or independent contractors. (Borello, supra, 48 Cal.3d at p. 345.) In answering this question, we began by explaining that the Workers’ Compensation Act distinguishes between covered employees and noncovered independent contractors based on “the common law ‘control-of-work’ test,” under which an employment relationship exists if “ ‘the person to whom service is rendered has the right to control the manner and means of accomplishing the result desired. . . .’ [Citations.]” (48 Cal.3d at p. 350.) We next held that, because “the ‘control’ test, applied rigidly and in isolation, is often of little use in evaluating the infinite variety of service arrangements,” it is appropriate to consider various “ ‘secondary’ indicia of the nature of a service relationship.” (Ibid.) We found the relevant secondary indicia in multiple sources. From our own decisions, we identified “ ‘the right to discharge at will, without cause,’ ” as a factor. (Ibid.) We then listed “[additional factors [that had] been derived principally from” the Restatement Second of Agency (sometimes, Restatement). (Borello, supra, at p. 351.) We then identified factors the Legislature had identified in Labor Code section 2750.5. (Borello, supra, at p. 351, fn. 5.) Finally, we identified factors the federal courts had adopted in applying the federal Fair Labor Standards Act of 1938 (29 U.S.C. § 201 et seq.; FLSA factors). (Borello, supra, at pp. 354-355.) All of these factors, we held, are relevant to determining whether someone is an employee under the workers’ compensation law. (48 Cal.3d at pp. 354-355.)
Consistent with Borello, in determining whether common questions of law or fact predominate, the trial court principally focused on Antelope Valley’s “right to control.” It found that the evidence the parties had submitted “demonstrates that there is no commonality regarding the right to control” and that “heavily individualized inquiries [will be] required to conduct the ‘control test.’ ” Regarding the secondary factors, the trial court found that “[s]ome carriers use helpers or substitutes”; “some carriers have multiple clients and customers; some have distinct occupation or delivery businesses; there is no commonality in the instrumentalities, tools, and place of work; *545carriers may or may not take advantage of chances to generate profits; and the length of time to perform services varies.” Contrary to what the majority opinion suggests (maj. opn., ante, at p. 534), in reaching these conclusions, the trial court considered several aspects of Antelope Valley’s contracts with its carriers, specifically the following: (1) some contracts included a “dock closing policy” while others did not; (2) the contracts did not require carriers to purchase mbber bands or bags from Antelope Valley; (3) the contracts stated that carriers may not put newspapers in containers that Antelope Valley has not approved; (4) the contracts provided for complaint charges; (5) the contracts provided some carriers with an option to redeliver newspapers to correct complaints, but that option was not available for certain routes or areas; and (6) the contracts provided that carriers could increase their profits by increasing the number of subscribers in their areas of delivery. The trial court also considered evidence — declarations and deposition testimony— showing how individual carriers actually performed their delivery duties, how Antelope Valley actually exercised control over the delivery process and its carriers as a group, and how Antelope Valley actually exercised such control with respect to particular carriers.
III. The Trial Court Erred in Making Its Ruling.
I agree with the majority that the trial court abused its discretion in making its mling and that remand is appropriate for further consideration of the certification motion. As noted above, in denying certification, the trial court found that the evidence “demonstrates that no commonality exists regarding the right to control.” However, the evidence plaintiffs submitted and principally relied on in support of their certification motion — including the form contracts between Antelope Valley and its carriers and the delivery instructions (known as “bundle tops”) that Antelope Valley typically prepared and provided to all carriers each day — shows that there is, in fact, some commonality in the proof regarding Antelope Valley’s right of control. Moreover, there surely is some commonality of proof regarding at least some of the secondary factors that are relevant under Borello to determine whether someone is an employee or independent contractor. Thus, in terms of proof, the trial court’s “no commonality” finding lacks support in the record and reflects insufficient consideration of the common proof plaintiffs submitted.
In addition, as we have explained, “ ‘that each [putative] class member might be required ultimately to justify an individual claim does not necessarily preclude maintenance of a class action.’ [Citation.] Predominance is a comparative concept, and ‘the necessity for class members to individually establish eligibility and damages does not mean individual fact questions predominate.’ [Citations.] Individual issues do not render class certification inappropriate so long as such issues may effectively be managed. [Citations.]” *546(Sav-On, supra, 34 Cal.4th at p. 334.) The “ultimate question” is whether “the issues [that] may be jointly tried, when compared with those requiring separate adjudication, are so numerous or substantial that the maintenance of a class action would be advantageous to the judicial process and to the litigants.” (Collins v. Rocha, supra, 7 Cal.3d at p. 238.) “The relevant comparison lies between the costs and benefits of adjudicating plaintiffs’ claims in a class action and the costs and benefits of proceeding by numerous separate actions — not between the complexity of a class suit that must accommodate some individualized inquiries and the absence of any remedial proceeding whatsoever.” (Sav-On, supra, at p. 339, fn. 10, italics omitted.) The record indicates that the trial court did not make the necessary comparison; it focused on the individualized proof it believed would be necessary regarding Antelope Valley’s right and actual exercise of control, and gave little or no consideration to the common proof plaintiffs submitted on these issues. By failing to make the legally required comparison, the trial court abused its discretion. I therefore agree we should affirm the Court of Appeal’s judgment and remand for additional consideration of the certification motion.
IV. The Majority’s Opinion.
Although I agree with the majority’s result, I question several aspects of its analysis. I begin with the fundamental rule that “[o]n appeal, we presume that a judgment or order of the trial court is correct, ‘ “[a]ll intendments and presumptions are indulged to support it on matters as to which the record is silent, and error must be affirmatively shown.” ’ [Citation.]” (People v. Giordano (2007) 42 Cal.4th 644, 666 [68 Cal.Rptr.3d 51, 170 P.3d 623].) Consistent with these principles, to the extent the trial court’s order is ambiguous, we must “resolve the ambiguity in favor of affirmance.” (Piscitelli v. Salesian Soc. (2008) 166 Cal.App.4th 1, 7, fn. 9 [82 Cal.Rptr.3d 139].) I find the majority’s approach, which generally seems to read the trial court’s ruling in the most unfavorable light, to be out of step with these well-established principles of appellate review.
More specifically, I find many of the numerous criticisms the majority levels at the trial court’s ruling to be off the mark. For example, I disagree that the trial court “ultimately rested” its order on variations “in the degree to which Antelope Valley exercised control over its carriers” and the circumstance that “the putative class as a whole was not subject to pervasive control as to the manner and means of delivering papers,” thus “los[ing] sight” of the relevant question “at the certification stage” (maj. opn., ante, at p. 534): “is there a common way to show Antelope Valley possessed essentially the same legal right of control with respect to each of its carriers” (id. at p. 533). As noted above, in finding insufficient commonality, the trial court expressly considered the extent to which the contracts showed either variations or *547uniformity in Antelope Valley’s “right to control” regarding several issues, including complaint charges, when carriers perform their work, use of unapproved containers, redelivery, and the carriers’ ability to increase profits. Indeed, the majority acknowledges elsewhere in its opinion that the trial court did, in fact, consider ways in which the contracts show “a uniform right of control, or uniform lack of right,” but it then criticizes the trial court for considering evidence that the parties’ actual course of conduct was different. (Maj. opn., ante, at p. 536.)
Contrary to the majority’s criticism, the trial court’s analysis was completely consistent with — indeed, was actually required by — Borello. There, we stressed that the right-to-control test “is not necessarily the decisive test” (Borello, supra, 48 Cal.3d at p. 351, fn. 5), that “common law principles are not dispositive of the employment relationship . . .” (id. at p. 352, fn. 6), that “[t]he nature of the work, and the overall arrangement between the parties, must be examined . . .” in addition to the right to control (id. at pp. 353-354), and that “[e]ach service arrangement must be evaluated on its facts, and the dispositive circumstances may vary from case to case” (id. at p. 354). Consistent with these statements, in finding as a matter of law that the agricultural laborers in Borello were employees, we explained that the evidence showed that the grower, “though purporting to relinquish supervision of the harvest work itself’ (id. at p. 355), actually “exercise[d] ‘pervasive control over the operation as a whole’ ” (id. at p. 356). Significantly, we expressly noted that, given this evidence of the grower’s actual exercise of control, a contractual provision purporting to give the laborers joint control over acceptable buyers was entitled to “little credence.” (Id. at p. 356, fn. 7.) Thus, in considering the parties’ actual course of conduct in addition to the contracts, the trial court here simply did what Borello required it to do. It also did what the record shows both plaintiffs and Antelope Valley urged it to do. Applying Borello at the certification stage, the trial court had to determine the extent to which there would be common proof regarding the “overall arrangement between” Antelope Valley and each of the putative class members. (Id. at p. 353.) The trial court’s ruling indicates that the trial court did precisely that.
Nor do I agree with the majority that the trial court, “by finding” that “variations in the actual exercise of control” were “sufficient to defeat certification, erroneously treated them as the legal equivalent of variations in the right to control.” (Maj. opn., ante, at p. 536.) The majority’s view appears to stem from its belief that the determination of whether the carriers were employees or independent contractors turns only on Antelope Valley’s right to control. As explained above, Borello establishes otherwise. So does another decision on which Borello extensively relied and which the majority cites: Tieberg v. Unemployment Ins. App. Bd. (1970) 2 Cal.3d 943, 946 [88 Cal.Rptr. 175, 471 P.2d 975]. (Borello, supra, 48 Cal.3d at pp. 349-351; maj. opn., *548ante, at p. 535.) There, we held that, in determining that certain television writers were employees and not independent contractors, the trial court had “improperly” declined to consider relevant secondary indicia, including the Restatement factors. (Tieberg, supra, at p. 946.) We also noted that “the terminology used in an agreement is not conclusive . . .,” even if it states that one party has “ ‘complete control of the services which the employees will render.’ ” (Id. at p. 952.) Nevertheless, we upheld the trial court’s determination because the trial court had relied not “solely upon” the alleged right to control under “the provisions of the contract,” but also on evidence that it had “in fact exercised control and direction over the writers.” (Ibid.) Thus, under Borello and its predecessors, evidence of an alleged employer’s “actual exercise of control” has independent significance, and “variations in the . . . exercise of control” are independently relevant to the certification question. (Maj. opn., ante, at p. 536.) Accordingly, there is no basis for the majority’s assumption that by resting its decision in part on “variations in the actual exercise of control,” the trial court was “treat[ing] them as the legal equivalent of variations in the right to control.” (Ibid.) Moreover, given Borello and Tieberg, the majority errs in stating that “how much control a hirer exercises” does not “matter[],” and that the only thing that “matters” is “how much control the hirer retains the right to exercise” (maj. opn., ante, at p. 533) and whether “there were variations in” Antelope Valley’s “underlying right to exercise” control over its carriers (maj. opn., ante, at p. 534, italics omitted).
I also do not entirely agree with the majority’s assertion that, “[a]t the certification stage, the importance of a form contract is not in what it says, but that the degree of control it spells out is uniform across the class.” (Maj. opn., ante, at p. 534.) As noted above, to assess predominance, a court “must determine whether the elements necessary to establish liability are susceptible of common proof or, if not, whether there are ways to manage effectively proof of any elements that may require individualized evidence. [Citation.]” (Brinker, supra, 53 Cal.4th at p. 1024.) Insofar as the terms of a form contract make clear that the alleged employer’s right of control is extensive, it is mdre likely that the elements necessary to establish liability will be susceptible of common proof and that there will be ways effectively to manage proof of elements that may require individualized evidence. Insofar as the terms of a form contract provide that the alleged employees retain extensive control over the details of their work, it is less likely that the elements necessary to establish liability will be susceptible of common proof and that there will be ways effectively to manage proof of elements that may require individualized evidence. Accordingly, what a form contract says may be of considerable importance in determining whether common issues predominate.
I also disagree we should limit our analysis to whether the trial court abused its discretion in applying the “common law test” for employment, and *549should ignore the FLSA factors Borello adopted. (Maj. opn., ante, at pp. 531, 532 & fn. 3.) Contrary to what the majority indicates, neither the trial court nor the Court of Appeal applied only “the common law test.” (Maj. opn., ante, at p. 531.) The trial court never used the phrase “common law” in either its ruling or during hearings on the certification motion. It did, however, consistently refer during the hearings to “the Borello factors” and “the criteria from the Borello case” and cite in its subsequent written ruling at least two of the FLSA factors Borello adopted: whether the carriers “use helpers or substitutes from time to time” and whether they can and do take action to “increase their profits” and “compensation.”1 (See Borello, 48 Cal.3d at p. 355 [relevant FLSA factors include “the alleged employee’s opportunity for profit or loss depending on his managerial skill” and his or her “employment of helpers”].) Consistent with this ruling, the Court of Appeal explained that Borello (1) adopted both the Restatement factors and the “six-factor test developed by other jurisdictions,” and (2) “cautioned that the individual factors — from the Restatement as well as the six-factor test — ‘ “cannot be applied mechanically as separate tests; they are intertwined and their weight depends often on particular combinations.” ’ [Citation.]” Nor did the Court of Appeal use the phrase “common law,” with a single exception: in quoting the passage of Borello that explains why “ ‘the “control” test,’ ” which derives from “ ‘common law tradition,’ ” “ ‘is often of little use in evaluating the infinite variety of service arrangements.’ ”
The lower courts’ application of all the Borello factors is consistent with the arguments the parties made below. In the briefs they filed in support of their certification motion and during argument on the motion, plaintiffs argued that “[n]ewspaper delivery is an integral part of [Antelope Valley’s] business” and that “the carriers perform an integral part of [Antelope Valley’s] newspaper business.”'“[W]hether the service rendered is an integral part of the alleged employer’s business” is one of the FLSA factors Borello adopted. (Borello, supra, 48 Cal.3d at p. 355.) Moreover, in their briefing, plaintiffs relied on the “factor” analysis the Court of Appeal used in Antelope Valley Press v. Poizner (2008) 162 Cal.App.4th 839 [75 Cal.Rptr.3d 887] (Poizner) to find that Antelope Valley’s carriers were employees rather than independent contractors. Poizner, which involved Antelope Valley’s workers’ compensation insurance premium, looked to all of the factors Borello cited, including the FLSA factors. (Poizner, at p. 853.) Thus, plaintiffs did not, as the majority asserts, “proceed[] below on the sole basis that they are employees under the common law.” (Maj. opn., ante, at p. 531.) Likewise, in *550its opposition to the certification motion, Antelope Valley relied on the same two FLSA factors the trial court cited in its written ruling: some carriers use “helpers or substitutes” to fulfill their contractual obligation and carriers have the “ability to generate profits or incur losses.” Thus, the record does not support limiting our analysis to the common law test for employment and ignoring the FLSA factors.2
Finally, I do not endorse the majority’s dicta regarding the “interplay” between “the predominance inquiry” and the Restatement factors in determining whether someone is an employee or an independent contractor. (Maj. opn., ante, at p. 538.) In light of the majority’s conclusion, that discussion is unnecessary. Substantively, it is also questionable in at least one respect. The majority asserts that certain Restatement “factors that might on their face seem to turn solely on the peculiarities of the parties’ particular arrangement, the Restatement intended to depend as well on general custom with respect to the nature of the work.” (Maj. opn., ante, at p. 538.) However, the Restatement comment the majority quotes in support of this assertion describes, not multiple “factors,” but only one factor: whether “ ‘the parties believe or disbelieve that the relation of master and servant exists ....’” (Maj. opn., ante, at pp. 538-539.) Moreover, the comment indicates, not that this factor depends in part “on general custom with respect to the nature of the work” (maj. opn., ante, at p. 538), but that a separate factor is “community custom in thinking that a kind of service ... is rendered by servants” (Rest.2d Agency, § 220, com. m, p. 492; see also id., com. h, p. 489 [listing as separate factors “the fact that the community regards those doing such work as servants” and “the belief by the parties that there is a master and servant relation”]).
For the preceding reasons, I concur in the judgment.
Like its final written ruling, the trial court’s written tentative ruling did not mention the “common law.” It did, however, state the court’s intention to deny the certification motion because “many” of the putative class members “will be found to be true independent contractors” because they do not “satisfy the Borello factors for determination of employee vs. independent contractor status.”
My conclusion reflects only fidelity to the record, not, as the majority asserts, any “assumption” on my part. (Maj. opn., ante, at p. 532, fn. 3.) In determining whether an order denying certification was error, an appellate court should apply “the theory on which plaintiffs pursued class certification.” (Fairbanks v. Farmers New World Life Ins. Co. (2011) 197 Cal.App.4th 544, 560 [128 Cal.Rptr.3d 888].)